Cardiac pulsatility and aortic compliance may result in aortic area and diameter changes throughout the cardiac cycle in the healthy and diseased human ascending, descending and abdominal aorta, as visualized by computed tomographic angiography (CTA), magnetic resonance angiography (MRA) and M-mode ultrasound [¹–⁴]. The presence of dimensional changes in the aortic root (composed of the aortic valve annulus, the site of attachment of the aortic valve leaflets; the sinuses of Valsalva and the sinotubular junction [⁵]) has been extensively studied over the past few years. No dynamic changes could be observed in the human aortic root in three different studies [⁶–⁸]. However, in dogs the commissural diameter of the aortic root increased by 12% between systole and diastole [⁹]. Furthermore, finite element model analysis [¹⁰] showed enlarged dimensions of the aortic root during the systolic phase of the cardiac cycle. These conflicting findings have resulted in an ongoing discussion about the dynamic morphological changes in the aortic root [¹¹, ¹²]. According to the observed dynamic changes at the level of the ascending, descending and abdominal aorta, we hypothesize that these changes are present at the level of the aortic root as well. However, even in the ascending, descending and abdominal aorta it is unknown in which phase of the cardiac cycle (systole or diastole) the dimensions are the largest.

The purpose of this study was to utilize ECG-gated multidetector row computed tomography (MDCT) and characterize the normal aortic root dimensions on an individual level and evaluate dynamic changes of these dimensions during systole and diastole.

We enrolled a total of 108 subjects (89 males, 19 females) with a mean age of 56.1 ± 12.5 years. The baseline characteristics, including a gender sub analysis, are demonstrated in Table 1.

The overall image quality of the 108 MDCT-scans used for this study was good. Scans with severe artifacts were excluded.

In the present study, ECG-gated MDCT (64- and 256 slice) was used to measure the dimensions of the aortic root (AV annulus, sinuses of Valsalva and STJ) in 108 healthy subjects. There was a significant difference between these three-dimensional measurements and the two-dimensional measurements, represented by the coronal measurement on MDCT. Other studies confirmed these results [⁶, ⁷].

With MDCT, previous studies did not show any significant difference between the systolic and diastolic dimensions of the aortic root. This unexpected finding might be caused by the method used to measure the dimensions. In these studies the dimensions during systole and diastole were measured in coronal and single oblique sagittal [⁶, ⁷] and in longitudinal reconstructions [⁸]. In our study, the axial reconstruction was used to measure the dimensions at different levels in the aortic root.

A comparison between the Dmax and Dmin on the axial image revealed a significant difference between both values that was variable between subjects. Seventy-six subjects (70.3%) had an oval shaped root (defined as a ≥3 mm difference between Dmax and Dmin [⁶]) during both systole and diastole. Two subjects (1.9%) had just an oval shaped annulus during systole and 19 subjects (17.6%) just during diastole. 11 subjects (10.2%) had a more circular shaped annulus. The fact that the aortic valve annulus is an oval shaped structure, rather than a circle, is confirmed by other studies [⁶, ⁷]. The difference in shape between systole and diastole implies a change in dimensions during the cardiac cycle.

Although many studies showed dynamic changes in the aorta diameter at different levels during the cardiac cycle, the presence of systolic and diastolic changes at the level of the aortic root were not observed in previous human studies by noninvasive imaging with MDCT [⁶–⁸]. Sonomicrometry studies in dogs [⁹] and finite element model analysis [¹⁰] did show enlargement of the aortic root dimensions during systole compared to diastole. There is an ongoing discussion [¹²] regarding the mechanism by which the aortic root dilates prior to opening of the aortic valve. Can this dilatation be explained by passive fluid dynamics or is it an active process and are there physiological or pathological processes that can influence this dilatation?

Angiography and echocardiography are helpful imaging modalities in the visualization of the aortic root. With the advent of transcatheter AV implantation procedures (TAVI) [¹⁶–¹⁸], the closed chest imaging becomes even more important, because of the difficulty with prosthesis sizing, accurate positioning of the prosthesis, the covering of the coronary ostia by the prosthesis [¹⁸] and even occlusion of the left coronary artery [¹⁷]. Currently used angiography and echocardiography are limited by their two-dimensional character [⁶–⁸]. MDCT can provide three-dimensional images with a high spatial resolution and give detailed information on the anatomy of the aortic root and the relation of the annulus with the coronary arteries [⁶]. The use of real-time 3D TEE to guide cardiac interventions has increased over the last few years. Direct planimetry of the aortic annular area by 3D TEE volumetric imaging showed the best agreement with MDCT as a ‘gold standard’ [¹⁹], although it still underestimated the MDCT planimetered areas by up to 9.6%, most likely due to a lower spatial resolution associated with 3D TEE volumetric imaging [¹⁹].

To demonstrate differences between systolic and diastolic dimensions, an individual comparison between both values was made. This comparison revealed significant dynamic changes in the aortic root dimensions. When the individual measurements were added to calculate the mean of the entire study population, this dynamic difference disappeared, as was the case in the other studies [⁶–⁸]. This can be explained by the highly variable distribution of the difference between systolic and diastolic dimensions, as was shown in the histogram. As expected, most of the individual subjects had larger systolic dimensions, compared to the diastolic dimensions, however, some had larger diastolic dimensions. According to these results we can conclude that there are dynamic changes in the aortic root of healthy subjects, however, they are very unpredictable.

With the limited available literature on this subject, an explanation of this unpredictability is challenging. Most of the studies describing the aortic root dimensions were carried out either by models or at autopsy [²⁰, ²¹]. Observations of the aortic root in the beating heart has been done both indirectly and directly. Indirect methods include X-ray techniques, with either contrast material (human study) [²²], radiopaque markers (animal studies in dogs and sheep) [⁹, ²³–²⁵] or sonomicrometry (animal studies in sheep) [¹¹, ²⁶] as well as echocardiography (animal study in dogs and a human study) [²⁷, ²⁸]. Direct observation of the aortic root in dogs can be achieved with cinematographic techniques (animal study in dogs), in which blood can be a disturbing factor [²⁹] or electromagnetic induction (animal study with dogs) [³⁰]. The studies performed in dogs and sheep with various imaging techniques all showed expansion of the aortic root during systole. A 16% radial displacement of the commissures was observed in an isolated aortic root model [²¹] and 12% in dogs [⁹]. The aortic root expansion starts prior to ejection and follows a precise order. Each root level reached maximal expansion during the first third of ejection [²⁶]. At end-diastole, the aortic root had a truncated cone shape and during systole it became more cylindrical in a sonomicrometry study in sheep [²⁶]. This result corresponds to our findings. However, Dagum et al. [²⁵] described an annular contraction during the ejection phase in sheep. In summary, there is an ongoing discussion with regard to the dynamic behaviour of the aortic root. The main limitation of the animal studies is the invasive, acute and open-chest nature of the models. Furthermore, the findings in dogs and sheep are not necessarily applicable to humans. No previous studies in humans proved pure systolic expansibility of the aortic root [⁶–⁸, ²⁸]. The lack of dynamic behaviour of the aortic root was attributed by the authors to the lack of elasticity in the aortic root in comparison to the much younger dogs and sheep used in the animal studies [⁸, ²⁸]. Howard et al. [³¹] has simulated stiffening of the aortic root and showed a disappearance of the radial expansion of the aortic root when stiffened.

The present study describesa non-invasive closed-chest method to measure the dynamic aortic root behaviour in the axial plane with MDCT. This study showed that there is a high variability between the different individuals, without any correlation with age, body length or body weight. The mechanism of the aortic root dynamics can only be partially understood with the available data and the knowledge from literature. Better understanding of the aortic root dynamic anatomy is a necessity. Because of the high individual variability, shown in this study, an individual non-invasive determination of the dynamic characteristics of the aortic root in each subject is necessary for adequate analysis. The method used by Van Prehn et al. [³²] could be an interesting method for segmentation of the different levels of the aortic root in the axial plane. With this method dynamic changes over the complete cardiac cycle in all directions can be analyzed in the thoracic aorta. In the future, this might support the cardiothoracic surgeons, cardiologists and engineers in their prosthesis selection, sizing and design.

In conclusion, the AV annulus is an oval, rather than a circular structure. Previous studies could not demonstrate dynamic changes in the aortic root, although they were present in the ascending, descending and abdominal aorta. This may be explained by the presence of a substantial variability in the dynamic changes of the aortic root dimensions during systole and diastole in healthy subjects. This study demonstrated a significant dynamic change in all subjects with a variable distribution between systole and diastole, due to the complex geometry of the aortic root. More research is needed to determine the exact shape changes in all directions throughout the entire cardiac cycle in both health and disease.

The authors gratefully acknowledge the support of The Netherlands Heart Foundation and the Smart Mix Program of the Netherlands Ministry of Economic Affairs and the Netherlands Ministry of Education, Culture and Science. Furthermore, the authors wish to thank Frederiek de Heer for her statistical assistance.

Conflict of interest None.

Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.