Is Antenatal Diagnosis of Coarctation of the Aorta Possible?
This article defines early gestational age at identification of asymmetry, the presence of a left sided superior vena cava, and a bicuspid aortic valve as strong predictors of a true coarctation abnormality. Further details can be found in the original article.
The original article is taken from a French journal and has been translated below. In the original article there are several graphs that can be seen. To view the original article click the link below.
Original Article (In French) - <Click Here>
Article Translated in English
The diagnosis of coarctation of the aorta during the prenatal period is difficult because the abnormality is not final in appearance until the postnatal period. During fetal life, we may only have suspicions based on indirect echocardiographic evidence. It has been known for over 20 years that an antenatal diagnosis of ventricular or outflow tract asymmetry, with a smaller left side, is a prognostic factor for coarctation, but it is not specific and may be associated with other conditions. There has been no report to date that seeks to precisely define the predictive factors for coarctation after birth.
Goal of the study:
To precisely identify antenatal factors predictive of coarctation of the aorta in our population.
Patients and Methods:
This is a retrospective study, encompassing a 4 year period (2001-2005), during which time 202 consecutive cases of fetuses were referred to L’Institut de Puericulture with a diagnosis of ventricular asymmetry with a smaller left side. The reason for consultation was the identification of ventricular asymmetry during a routine 2nd or 3rd trimester study. Only overt ventricular asymmetry with a right ventricle to left ventricle ratio of >1.5 were included. Cases with associated congenital cardiac malformations (with the exception of ventricular septal defects) were excluded from the study population.
Of these 202 cases, 17 were terminated due to associated structural abnormalities and/or fetal aneuploidy. In 22 cases, the patients were lost to follow up with no further antenatal studies and no information regarding pregnancy outcome. The majority of second trimester studies performed underwent a third trimester follow up. Of the 163 patients born alive, 147 underwent postnatal echocardiography either in our center or near their location of birth. The remaining 16 newborns were not examined prenatally in our center but had normal postnatal echocardiograms.
The diagnosis of ventricular asymmetry was confirmed on average at 31 weeks GA. In 39 fetuses (24%), this was made in the second trimester, meaning prior to 28 weeks GA. Of the 147 fetuses seen pre- and post-natally, 31 developed coarctation. The 16 fetuses not seen prenatally all had normal postnatal echocardiograms excluding coarctation. Tables 1 and 2 summarize the clinical and echocardiographic characteristics of the entire population during the 2nd and 3rd trimesters. In the group of patients that evolved to a diagnosis of postnatal coarcation, the gestational age of diagnosis was significantly different than the normal outcome group, with a median gestational age of diagnosis of 26 vs 34 wks GA. The arterial outflow tract asymmetry was significantly more pronounced in the coarctation group, with a pulmonary artery to aorta ratio (PA/AO) of approximately 1.8. This difference was confirmed in the 3rd trimester with the same ratio obtained. There was no evidence of significant growth or change in the PA/AO ratio in the 39 (2nd tri) fetuses diagnosed prenatally with asymmetry. Similarly, for the population of fetuses without coarctation, the PA/AO ratio was also stable from the 2nd to the 3rd trimester. The size of the mitral valve annulus was not significantly different between the groups with and without coarctation. Almost half of the cases of 2nd trimester ventricular asymmetry evolved towards a diagnosis of coarctation, in contrast to only 10% of the cases of 3rd trimester asymmetry (Table III). Slightly less than half of the coarctation cases demonstrated a LSVC, and greater than ¾ of the cases had a bicuspid aortic valve. Table VI shows the positive and negative predictive validity of factors associated with coarctation. The combination of a LSVC and an early (2nd tri) diagnosis of ventricular asymmetry has a greater than 70% likelihood of having a coarctation.
Since the beginning of echocardiography of congenital heart abnormalities, predicting the development of coarcation postnatally has remained a difficult challenge; however, it is also indispensible as it would allow for a significant reduction in morbidity and mortality as a result of this lesion. The increased frequency of associated extra-cardiac and chromosome abnormalities (up to 40% in certain series) adds to the diagnosis and in many cases would justify the use of fetal karyotyping. In Paladini’s series, the most common aneuploidies were Monosomy X, Trisomy 21, 18 and 13. Extra-cardiac abnormalities included cerebral malformations (Dandy-Walker, agenesis of the corpus callosum, holoprosencephaly), GI anomalies (esophageal and anal atresia), renal anomalies (renal agenesis, pyelectasis or hydronephrosis, hypospadias), and various skeletal anomalies or dysplasias. In our series, approximately 10% of the cases were terminated for associated anomalies or aneuploidy. There is little follow up on these patients available.
At the end of the 1980’s, the first papers describing the prenatal echocardiographic characteristics of fetuses at risk for coarctation appeared. The finding of ventricular asymmetry seems to have good predictive value for the development of coarctation, but lacks sensitivity and does not allow the subsequent detection of all coarctations. The specificity was equally poor in our series, with a false positive rate of approximately 80%. Ventricular asymmetry is perhaps too vague a criterion, and the works of Sharland (4) add to the confusion by concluding that ventricular or arterial (outflow) asymmetry does not allow for the prediction of coarctation. The study of David et al (5) brings an important element to the interpretation of ventricular asymmetry by introducing the concept that early (2nd tri) asymmetry specifically does allow for the prediction of subsequent coarctation. This is readily apparent in our series, where approximately ½ of the fetuses with 2nd tri asymmetry evolved towards coarctation, in contrast to only 10% with 3rd trimester asymmetry. With the ongoing development and improvement of ultrasound machines, the mid-90’s saw growing interest in previously poorly visualized structures, such as the ascending aorta, the pulmonary trunk and the distal arch, and transverse arch/isthmus. The distal arch was studied in a series by Hornberger (6). Severe hypoplasia of the transverse arch (<3rd percentile) was found in 60% of fetuses with an isthmic anomaly, specifically coarctation. However, this series includes only 20 patients, almost all with severe associated congenital cardiac abnormalities as well as extracardiac or chromosomal abnormalities, which represents a very specific population and one that is substantially different from our own. The study of the distal arch specifically was not possible in all fetuses in the series due to technical and imaging limitations. It is easier to study the origin of the great arteries and their respective diameters. In our series, as previously described, the relationship (ratio) between the pulmonary artery and aorta (PA/AO) is significantly elevated in cases of coarctation, either 2nd or 3rd trimester. Approximately 2/3 of fetuses followed with a PA/AO ratio > or = to 1.6 had a coarctation after birth. The ratio or relationship is relatively stable over the course of the 2nd and 3rd trimesters, and does not allow further identification of those destined to develop coarctation. Bicuspid aortic valve is a very strong association in favor of coarctation (9/10) but this remains almost an exclusively postnatal diagnosis. With improvements in imaging and the use of 3D/4D technology, the prenatal diagnosis of a bicuspid valve and therefore fetuses at greater risk may be possible. Notwithstanding this association of a bicuspid valve with coarcation, it should be remembered that this is rarely an anomaly seen in isolation and if identified in a newborn at risk, ongoing surveillance until ductal closure is prudent. A left sided SVC, despite having minimal clinical significance, is frequently associated with congenital heart disease and in particular coarctations. The hypothesis is that a reduction in flow across the mitral orifice, which is partially obstructed by the LSVC, is responsible for a reduction in growth of left sided structures as a result of decreased flow. The presence of a LSVC in a fetus with an early diagnosis of ventricular asymmetry allows for an accurate prediction of coarctation in 70% of cases.
The accurate prenatal prediction of coarcation of the aorta is possible in 2/3 of cases of early ventricular asymmetry, due to the presence of associated additional findings such as a LSVC. However in 1/3 of cases, we must accept that ventricular asymmetry may be the only prenatal finding. This is inherently problematic as ventricular asymmetry alone is poorly specific for coarctation of the aorta. Basing follow up on this single criterion will amount to excessive surveillance in over 80% of cases (including prenatal surveillance, neonatal surveillance), is a source of anguish for parents, and is costly to society. Technical improvements in the coming years may allow particular surveillance for issues like bicuspid aortic valves, as well as assessing the morphometry and Doppler studies of the distal arch. A larger series of patients exhibiting specifically 2nd trimester ventricular asymmetry would also yield more information.
Stos B et al
Arch Mal Coeur Vaiss 2007 Mai 100 (5), 438-3
Supplemental Images to Article:
The image below demonstrates ventricular disproportion, with a large right vs left ventricle. This disproportion is usually felt to be legitimate when the ratio of ventricular diameters (measured just below the AV valves) is greater than 1.5. Of note, and purely sematic, one should not use the term ‘ventricular discordancy’ to describe this difference. This term is reserved for a discordance between ventricular location and morphology- for instance, when a left (or posteriorly) located ventricle has characteristics of a right ventricle, such as a moderator band and a septal papillary muscle attachment.
The next picture demonstrates the size difference between pulmonary and aortic diameters at the level of the 3 vessel view- also a feature of coartctations. The ratio of >1.8 is significant for underlying coarctation
Both pictures below demonstrate narrowing (coarctation) at the aortic isthmus. The second, as denoted by the red arrows, shows what is referred to as a ‘shelf’ which appears as a horizontal (compared to longitudinal orientation of the descending aorta) abruptness to the vessel wall just beyond the narrowed isthmus
3 vessel view demonstrating a left sided superior vena cava, which carries a strong predictive value when seen in conjunction with a suspected coarctation
Finally, this is a pulmonary valve, but the concept is the same. A narrowed region of interest across a volume of the valve, when rendered, can be helpful in demonstrating bicuspid from tricuspid valves. Bicuspid valves can be further subdivided into one that appears preliminarily tricuspid but with 2 fused leaflets, and a true bicuspid valve demonstrating a midline raphe where the valve leaflets meet. A very magnified view of the aortic valve should be obtained as a volume in cases where we are suspicious- in as close an orientation to the above noted image as possible.