How to measure the aorta in the setting of genetic aortic disease

Definition of aortic dilatation

In the context of genetic diseases of the aorta, a correct definition of aortic pathology is needed both in terms of diagnosis and treatment. When considering further work-up for genetic causes, it is important to define aortic dilatation and -aneurysm. Aortic dilatation in most genetic aortic disease entities is typically located at the root, at the level of the sinuses of Valsalva and is defined as a measured aortic diameter above the 95% confidence interval of the normal distribution in a large reference population. (1) (2).

To correlate measured aortic diameters with normal values, age- and body surface area (BSA)-related nomograms for upper limit of normal (ULN) or equations for z-score calculation should be used. The z-score is a standard numerical representation on how much a specific value defers from the average in a normalized distribution. A z-score of 0 means that the studied value coincides with the mean. Normal z-scores range from -2 to +2 standard deviations. Positive z-scores indicate that the value is above average while negative z-scores indicate that the value is below average. The z-score can be calculated using the formula z=(x- µ)/ σ  were x is the studied value, µ is the mean and σ the standard deviation. A z-score ≥2 indicates a value exceeding two standard deviations (the 95% confidence interval).

Z-scores correlate with percentiles. Percentiles show the ranking of a value in a set of 100 equal parts. A z-score of 0 corresponds to the 50th percentile while z-scores of 2 and 3 correspond to the 97,7^th and 99,9^th percentile respectively. The upper limit of normal (ULN) is usually set at the 95^th percentile.

The first nomograms for normalizing aortic root diameters (at the level of the sinuses of Valsalva and the supra-aortic ridge) with 2D echocardiography were established by Roman and colleagues in 1989 (3) . In this study 135 adults and 52 children were included and the ULN was defined based on BSA. These nomograms, which are widely used in clinical practice and adopted in guidelines, made a subdivision in three different age categories, leading to jumps in diameter prediction when transitioning from one age category to the next one, as illustrated in figure 1.

This problem was resolved in the updated nomograms by the same group for aortic root ULN

prediction, considering age as a continuous variable(4). Meanwhile, several other aortic nomograms have been developed. Nomograms are derived from regression equations including age, sex and bsa as the most common variables. Figure 2 illustrates the importance of taking these factors into account and an illustration of nomograms in which obtained values can be correlated with expected values is provided in Figure 3 (Figure 2&3).

Table 1: Overview of the different published nomograms for z-score calculation

As shown in table 1, a major limitation in using these nomograms, is that most groups have limited their calculation to either pediatric or adult populations. More recently z-score and ULN formulas further extending the age range included both children and adults allowing for a unified interpretation throughout growth and aging. This paper also provides normative data on the tubular ascending aorta (5).

In addition to matching for age, sex and BSA, the method used to measure the aorta is equally important when comparing and calculating z-scores. This is especially true in the pediatric population where different ways of measuring the aorta are used and multiple nomograms exist. An illustration of this problem is given in Figure 4.

Limitations of the z-score equations and alternative calculations

In addition to the pitfalls mentioned above, calculation of z-scores are limited by the fact that not all ethnic groups are equally represented in the nomograms (most of the included patients were Caucasian) and the fact that over- or underweight can lead to an over or underestimation of the z-score as illustrated in Figure 5 (11)

Furthermore, the z-score obtained by different nomograms, even when appropriately used, can lead to a different z-score and subsequently different clinical interpretation. This was shown by Rutten and colleagues in a study of children with Marfan syndrome where the percentage of children with a z-score ≥ 2 changed depending on the nomogram used (35% using Lopez et al. versus 20% using Pettersen et al.) (12)

Current practice of state-of-the-art measurement of the aorta

Transthoracic echocardiographic views

Echocardiographic images are obtained from the parasternal long-axis view visualizing the aortic root and the proximal ascending aorta. Different transducer positions are often needed to optimize the images: a high left parasternal view located one or two intercostal spaces superior to the standard location may be required. Alternatively, a high right parasternal view in a right lateral decubitus position may better display the entire proximal aorta. 

How to measure

Measurements of the aortic root and tubular ascending aorta are taken at different levels: (1) the aortic valve annulus (defined as the hinge points of the aortic leaflets); (2) the maximal diameter at the sinuses of Valsalva; (3) the sinotubular junction (transition between the sinuses of Valsalva and the tubular portion of the ascending aorta); and (4) the ascending aorta at the level of the right pulmonary artery. In some patients aortic root dilatation is asymmetric. This can for instance be the case in patients with an underlying connective tissue disorder and in patients with a bicuspid aortic valve (especially patients with right and non-coronary cusp fusion). Dimensions in such cases can be inaccurate. The transverse plane of the aorta (parasternal short-axis view) may be of help to reveal this asymmetry, but also given the low lateral resolution, cross-sectional imaging with CMR or CTA is needed for accurate measurement (13).

When measuring aortic diameters, it is important to obtain the maximum diameter recorded perpendicular to the long axis of the vessel in that view. According to the current adult quantification guidelines all aortic measurements should be made at end-diastole using the leading-to-leading edge convention (DLE), except for annular measurements which should be performed at mid-systole from inner to inner edge (SIE) (14) (Figure 6). Pediatric guidelines on the other hand advise to measure all root and ascending aortic dimensions from SIE (15). These different recommendations can be challenging particularly in the transition of care from pediatric to adult cardiology departments. However, when comparing the two conventions only a non-significant underestimation of the diameter with the SIE compared to the DLE convention can be observed - indicating that both conventions can be interchangeably used in the adolescent age group(16) (17). The main reasons why adult guidelines recommend the DLE convention are (1) the fact that available normative data for aortic dimensions as well as recommendation thresholds for elective root and/or ascending aorta replacement are mostly based on the DLE method (3) (18) (2) (19)  – use of another method may adversely affect patient prognosis by delaying surgery and (2) measuring in diastole provides better reproducibility due to stable hemodynamics. M-mode measurements, as performed in earlier days, are no longer considered adequate for aortic root measurements due to motion of the heart during the cardiac cycle and changes in M-mode cursor location relative to the maximum diameter of the sinuses of Valsalva (14). This can result in a systematic underestimation (by up to 2 mm) of aortic diameter by M-mode in comparison with the 2D measurements of aortic diameter(3).

Comparison of methods for aortic measurements and dimensions obtained by CTA, CMR versus TTE

Differences in diameters obtained by the different imaging modalities can be expected for several reasons, (1) CTA and CMR show all dimensions of the aorta whereas TTE is inherently limited to its 2 dimensional view on the aorta which may not necessarily reflect the maximum aortic root diameter especially not in asymmetric aortic roots (20) and (2) there is no uniform method for measuring the aortic root and tubular ascending aorta among TTE, CTA and CMR. TTE uses the leading edge-to-leading edge convention, whereas CTA and CMR use the inner edge-to-inner edge or outer edge-to-outer edge convention(14). In CTA and CMR imaging, aortic measurements should be obtained on ECG-gated images, in end-diastole, using the double oblique methodology (21) (Figure 7).

For the measurement of the sinus of Valsalva three methods have been proposed: sinus-to-commissure (bisecting cuts), center sinus-to-sinus (off-center cuts) or maximum sinus-to-sinus (off-center cuts) measurement assessed in diastole, inner-to-inner edge (I-I) in the short axis view (figure 8). It remains unclear which of these methods is of most clinical relevance.  Nevertheless it is clear that the same method should consistently be used during the follow-up of each individual patient (22). The dimensions obtained by the maximum I-I sinus-to-sinus method dimensions are significantly greater compared to the other two methods(21, 23, 24). This method furthermore yields the best agreement with DLE TTE measurements, and particularly the maximum I-I sinus-to-sinus measurements of the right-to-non-coronary cups most closely agrees when assessed by CTA(23). Similarly Rodriguez-Palomares et al. showed that the diameter by DLE TTE correlated best with the I-I right-to-non-coronary cusp diameter determined by both CTA and CMR – of note the center sinus-to-sinus method was used for sinus-to-sinus measurements(21). Despite the close agreement there still exists an underestimation of diameters by TTE of about 2 to 3mm(21). Rodriguez-Palomares and colleagues demonstrated a minimal difference of 0.2 to 1.2mm (mid sinus-sinus technique used), whereas Frazao and colleagues found a 1.4 to 3.5mm difference (max sinus­­-sinus technique used) (25). Figure 9 summarizes the recommendations for measurements of the sinuses of Valsalva and tubular ascending aorta with different imaging modalities.

Conclusion

The importance of correct measurement of the aorta in the context of hereditary diseases of the aorta cannot be underestimated. In a diagnostic setting, z-scores correcting for age, BSA and gender are usually used and correct reference values should be used. To determine indications for surgery, the absolute diameter of the aorta is usually used and, here too, the correct technique and measurement method should be used.

In this paper we have focused on measurements of the aortic root - of course the distal aorta cannot be ignored and additional parameters such as arterial tortuosity and functional characteristics such as stiffness can also be used to better assess the diagnosis and risk.