New Screening Tool for Aortic Root Dilation in Children with Marfan Syndrome and Marfan-Like Disorders

New Screening Tool for Aortic Root Dilation in Children with Marfan Syndrome and Marfan-Like Disorders.

Wozniak-Mielczarek L, Sabiniewicz R, Nowak R, Gilis-Malinowska N, Osowicka M, Mielczarek M.

Pediatr Cardiol. 2020 Jan 31. doi: 10.1007/s00246-020-02307-0. [Epub ahead of print]

PMID: 32006082

Similar articles

Select item 20301322


Take Home Points:

  • Aortic root ratio may be another helpful screening measurement for aortic root dilation in patients with suspected connective tissue disorders
  • This study did not assess normal patients or patients with other etiologies for aortic root dilation (such as BAV)
  • Whether cumulative changes such as rate of change or significant change would improve morbidity/mortality or be useful for surgical intervention remains to be seen

Commentary from Dr. Jared Hershenson (Greater Washington DC), section editor of Pediatric Cardiology Journal Watch:  Similar to nearly all echo measurements in children, the aortic diameter is affected by a patient’s age, sex, and BSA. Z-scores are used to allow for the necessary adjustments in order to determine if something is within the normal range. There are currently 3 main z-score nomograms used. Based on a premise that z-score calculations are time consuming and impractical, this study aimed to introduce a simple screening method to identify aortic root dilation. The authors retrospectively analyzed 193 patients with Marfan Syndrome (MFS) or Marfan-like disorders (including Ehlers-Danlos, Loeys-Dietz, ectopia lentis, neonatal Marfan, MASS phenotype, and Marfanoid habitus). Marfanoid habitus was defined as a constellation of symptoms similar to MFS but did not fully meet Ghent criteria. The mean age was 12 years. The aortic root (sinuses of Valsalva only) was measured in the parasternal long axis view using both the leading edge in end-diastole and the inner edge in mid-systole. The aortic root ratio (ARr) was calculated as the aortic root diameter (mm) divided by the height (cm) multiplied by 100. This was compared to the 2 measurement techniques and z-scores using all 3 nomograms.


28-31 % of the patients had aortic root dilation as expressed by z-score (z > 2) depending on which nomogram was used. 11 patients (5.79%) had inconsistent results (dilated vs. non-dilated) when comparing z-score nomograms (Gautier vs. Cantinotti only). The mean value for ARr was 18.1 vs. 17.8 for the leading edge vs. inner edge measurements. The optimal cut-off for ARr was > 18.7 with resulted in a sensitivity of 88-100% and specificity of 94-98% (Table 7).

The authors further assessed the subgroup in whom the ARr failed to identify aortic root dilation. They found the results were inconsistent using the 3 different nomograms. Notably, a detailed table shows that all of the z-scores of this group were either just below or just above 2. ROC curves showed an ARr > 18 yielded a sensitivity of 100% for the Petterson and Cantinotti nomograms and > 18.7 for the Gautier nomogram. Tables 10 and 11 show the high values of PPV, NPV and accuracy of ARr using the various cut-offs.

The authors make a reasonable argument that extremely thin or obese patients may have inaccurate z-scores, and based on previous data that height is more important than BSA for aortic root diameter, the ARr may be useful. There are however a few major caveats to this study. First, ARr only assess the aortic root at the sinuses. It will therefore not be useful in those with significant annular dilation or ascending aorta dilation, and we know that surgical decision making often entails valve sparing vs. non-sparing aortic root replacement. Additionally, the Marfan-like subgroup was large and may not represent a normal population for which this measurement has not been validated. They also did not include a BAV/aortic root dilation group, or other groups with CHD such as TOF or d-TGA s/p ASO, which may have a different form of aortic root dilation than the genetic connective tissue disorder group. Finally, the claim that nomograms are difficult to use may be a bit spurious, given that many reporting software and measurement packages on the echo equipment may automatically calculate z-scores. It is interesting to note that there may be variation depending on the nomogram used, so echo readers should keep this in mind based on the specific software algorithms used. However, most of these patients would have routine follow up, so a “missed” diagnosis when the z-score is 1.9 is probably not too clinically relevant. Since this was a one-time measurement study, further research on the ARr over time would be necessary, since rate of change is also a key factor in decision making.