Looking For Fetal Heart Disease - SD

Introduction to Fetal Heart Screening

Hello, I'm Dr. Joshua Coppel, and I'm at the Leading Edge conference in Atlantic City, New Jersey, where I'll be talking this afternoon about some important information in screening for fetal heart abnormalities, the most common form of major congenital anomaly that's found in the United States, and in fact, around the world.

We'll be talking about some of the things you need to look for when you're doing routine obstetrical ultrasounds in order to identify fetal heart problems.

This talk is about looking for fetal heart abnormalities, and it's intended to be a really practical approach to looking at the fetal heart in the course of routine everyday prenatal scans.

I'm not here today to talk about complicated problems, to talk about how to do a formal fetal echocardiogram or to show a bunch of scary complicated images of hearts to show how great we are at finding heart problems.

Really, I want to be very practical, because that's what goes on every day in people's practices.

Importance of Looking at the Heart

The importance of looking at the heart is shown in the next few slides.

And while this is old data, the numbers have been very stable in the United States for several decades.

The rate of death for infants up to one year due to congenital abnormalities is the number one killer of babies up to one year.

It's more common than prematurity, low birth weight, things that we spend a lot of time trying to detect and prevent.

It's more common than SIDS, more common than any of the other things listed on this table.

In graphic form, more recent data from 2005, we see that congenital abnormalities remain the number one cause of death up to one year after birth.

Within the category of congenital abnormalities, this pie chart shows that about a third are due to cardiovascular abnormalities.

And again, while this data is old, numbers have not changed substantially in the last 20 years.

Structured Approach to the Four-Chamber View

When we look at the heart as part of a full echocardiogram, there are a whole series of views that we need to obtain in order to be sure that the segmental anatomy of the heart is normal.

But that's really not practical for everyday scanning as goes on in people's offices.

We need to have a structured approach to looking at the heart, and the best way to do it is to start by looking at the four chambers of the heart.

That's not as easy perhaps as it sounds because the bullet points on this slide list all of the various components of four chamber view that you need to keep in mind as you're looking at it.

It sounds simplistic perhaps to say the heart should be on the left side of the chest, but the heart should be on the left side of the chest.

And if it's on the right, even if the stomach is also on the right, that can represent a problem.

So you need to know where the head is, where the breech is, the spine, the stomach, and the heart, and be sure you know the sidedness of the baby.

The atria should be about equal to each other in sizes, as should the ventricles normally before birth.

There's a little more flow through the right side of the heart than to the left side.

So you can see the right atrium and ventricle looking a little plumper than the left atrium and ventricle, but they should be pretty close to each other in size.

The left atrium is the posterior most chamber, and the foramen ovale flap is usually visible in the left atrium.

When you look at the atrioventricular valves, you wanna see that the tricuspid valve is a little closer to the apex of the heart than the mitral valve.

They shouldn't line up straight across in the heart.

You should be able to see the interventricular septum and that it's intact.

You should see the moderator band in the right ventricle.

The heart should have an axis that's about 45 degrees off line from the center towards the left.

The normal range is 30 to 60 degrees, and the heart should occupy about a third of the chest area.

You should be able to put two more hearts inside the chest when you look at a cross-sectional view.

Normal Four-Chamber Views

Now, over the next number of minutes as we go through some possible abnormalities, I'll be showing you examples that relate to all of these things.

But let's start with some normal images.

We'll reminder, just look at the stomach, and here we have a normal four chamber view of the heart.

We see the descending aorta in cross section.

The spine is off on the left side of the screen.

The left atrium and the right atrium are about equal in sizes.

The left ventricle and right ventricle are about equal in sizes.

We see a little bit of the foramen ovale flap in the left atrium.

The mitral valve and tricuspid valve insertions are seen.

We see a little bit of the moderator band up at the apex of the right ventricle.

This is cut off, so it's hard to see that this only occupies a third of the chest, but we'll be able to see that better in some subsequent images.

Here's a long axis four chamber view, so-called because the septum is sitting horizontally on the screen.

Here, we've got a better sense of the whole circumference of the chest, and I think it's pretty clear that you could put at least two more hearts inside this chest.

The value of this particular image, a couple of values.

One, you can see the whole length of the interventricular septum.

It looks great.

Two, you can see the insertion of the tricuspid valve and of the mitral valve and the tricuspid valve in the upper two uppermost of the ventricles here, the tricuspid valve is a little closer to the apex of the heart than is the insertion of the mitral valve.

Then chambers on the right side about equal in size to the chambers on the left side.

Nice looking four chamber view.

If you look instead from the apex of the heart, you can see some things better and some things worse.

You get a better view of the moderator band in this particular view, but you're not able to see the valves or their insertion on the septum quite as well.

And you certainly can't see the upper ventricular septum as well.

That's because looking at this septum, which is aligned vertically on the screen or in parallel to our ultrasound beam, relies on lateral resolution of the ultrasound.

And we see better with axial resolution.

The previous image showed the heart in a position that where we were able to see using axial resolution, the sides of the ventricular septum.

But here it relies on lateral resolution.

We just can't see it as well.

Still, we can see that the chambers are about equal in sizes on the right and the left side.

The about one third of the chest area occupied by the heart.

All the other normal things about this are present.

As patients get heavier, it can get harder and harder to see the heart, particularly when the position isn't optimal.

Here's a patient, a diabetic who came in for a fetal echocardiogram, and this is from the midline of mom's abdomen.

A four chamber view of the heart.

I think you might agree that it's a heart, but whether it's a normal four chamber view, I certainly can't tell that there are probably four chambers there, but I can't say much about the ventricular septum, particularly when we moved our transducer way over on the patient's right side.

So we came around and tried to get at the fetal heart from the side.

We were able to get a view where the septum appeared intact, and we were able to see the insertion of the valves look fine.

The chambers look fine.

So you need to be creative and think about where you wanna be on the baby's chest and how you need to move on mom's abdomen to get to that position to optimize the image.

So those are some things about normal four chamber views.

Abnormal Findings

What's abnormal?

Well, it's anything that's not normal.

Anything that disturbs you, that makes you think that things just don't quite look the way they should, should really prompt a more detailed look at the heart.

The heart could be on the right, the chambers could be unequal in sizes.

We could have funny insertion of the valves, an obvious hole, anything at all that gets you concerned.

So let's look at some examples.

First, we'll start with the most common things, ventricular septal defects, and here's a fetus with an atrioventricular septal defect.

Also called an AV canal defect, or an endocardial cushion defect.

I've got two pictures here taken in systole and diastole because they illustrate different things.

We'll start with the left panel, the image in systole, and you can see that the atrioventricular valves are lined up linearly, just a straight line across there towards the top of the ventricular septum.

We don't see much of a ventricular septal defect in this panel, although we can see the atrial septal defect here.

When we're in diastole and the valves open, it's easy to see that there's a very large defect with loss of both atrial and ventricular tissue.

But the key picture to think about is this one over here on the left, where we have the linear appearance of the atrioventricular valves, because that can be your main clue to the presence of an atrioventricular septal defect.

This is another fetus with an atrioventricular septal defect showing the linear appearance of the atrioventricular valves.

Some of these fetuses with AV septal defects will have regurgitation, and that's one of the things that I particularly look for when I'm seeing them, because severe regurgitation, particularly when it's high velocity and holosystolic, it is illustrated in this picture, can lead to hydrops, and that can be very bad for the fetus combination of structural heart disease and hydrops.

So I'll always look at that and be watching that during the course of the pregnancy.

It drives a little bit how often we see these patients and how we counsel them.

Remember that all the VSDs that we see even on four chamber views are not necessarily atrioventricular septal defects.

There can be muscular or other VSDs that are visible.

Here's a very large ventricular septal defect.

I know this is an artifact that a little bit of brightness at the top of the septum there is what we see when there's actually a gap in tissue.

So this is a real finding, but it wasn't an atrioventricular septal defect.

They were two separate AV valves and a normal atrial septum.

We might also see that something looks too big, and here's an example of a fetus with something that's too big.

Actually, two things that are too big.

The right side of the heart is towards the top here, the left side of the heart.

We've got the left atrium close to the spine, we've got the left ventricle, and clearly the left side of the heart is smaller than the right side of the heart, or the way I guess I should put it, is that the right side is big compared to the left side.

And in fact, most of this is made up of a very large right atrium due to a dysplastic tricuspid valve that's allowing a great deal of valve regurgitation.

But the main thing to think about here as the frontline person doing the scan is when we look at the whole chest here, it's almost all heart.

There's only a little bit of lung and something's wrong with a heart to make it this big, something too small.

Now, while this is a fetus that's labeled small left ventricle, in fact, as if you looked at a panorama of this fetus, you would probably think that the right side is big as well as the left side being small, and that's what happens to fetuses that have left ventricular obstructions.

The systemic venous return usually stays on the right side.

The umbilical venous return is supposed to cross the foramen ovale to the left side, but it can't, if there isn't enough exit of blood from the left ventricle and pulmonary venous return may not be able to fill the left ventricle and have to go retrograde across the foramen ovale.

So the right side ends up being dilated because of the extra flow through it.

This is a fetus that turned out to have a coarctation of the aorta.

Clearly, there's a big disproportion in the sizes of the atria as well as the sizes of the ventricles.

Here's another extreme example of a fetus very early.

This is about 18 weeks, and this fetus had a severe left ventricular outflow obstruction resulting in hypoplastic left heart with fibroelastosis of the ventricular septum and of the free wall of the ventricle in compensation.

For that, the right side appears dilated because again, all of the flow that would've gone through the left side is now added to the flow on the right side.

Here's an even more extreme example.

At the very top of that arrow is a tiny black dot, and that's all that's left of the left ventricle and the fetus with a hypoplastic left heart.

Now, at first blush, first glance, you might say that you think this is a AV septal defect and that this is a common atrioventricular valve with this being the ventricular septum.

In fact, I know it's not because the atrial septum is over here and it should line up with the ventricular septum.

That can't be the ventricular septum.

So we have a small left atrium and an enlarged right atrium and right ventricle, barely visible left ventricle.

Here's another fetus with a small left ventricle, again due to a left ventricular outflow tract obstruction.

Key points on this particular image besides the disproportion in sizes is to note that the left ventricle goes all the way to the apex of the heart and is contracting well.

So there's still a chance that this will be a fetus that can have a two ventricle repair after birth.

It has not developed scarring down of the left ventricular chamber.

Some fetuses do develop that scarred appearance to the left ventricular chamber.

This still picture shows kind of a globular echogenic appearance to the left ventricle.

While it's not terribly small, it's not going to grow anymore during the course of the pregnancy.

So these fetuses will end up with hypoplastic left heart.

The right side tends to behave a little bit differently.

And here's a fetus with a hypertrophic kind of muscle bound right ventricle.

Here you see an enlarged right atrium.

You don't tend to see an enlarged left atrium with an obstructed left ventricular outflow, but you do get an enlarged right atrium when the right ventricular outflow is obstructed.

Whether that's pulmonic stenosis or anything else going on distal to the valve, you tend to get some tricuspid regurgitation and enlargement of the right ventricle.

So the two sides behave a little bit differently, and the clue to a right sided obstruction might be a big right side.

Here's a combination of things with maybe a little bit enlargement of the right atrium.

This is another case of pulmonic stenosis, muscle bound right ventricle dilated, ballooning foramen ovale flap into the left atrium because all of the systemic venous return is having to go across the foramen ovale and along with the umbilical venous return over to the left side of the heart.

This is a normal four chamber view in the fetus that was referred to us because the mother had pulmonic stenosis herself.

And the importance of this view is just as a reminder that you shouldn't do a screening test.

And the four chamber view is a screening test.

You shouldn't do a screening test in a patient who is at high risk.

If the patient's at high risk, you wanna get a diagnostic test.

And it was only through a fetal echocardiogram where we looked at the pulmonic valve and were able to identify that there was high velocity across the pulmonic valve to make that particular diagnosis.

Situs and Axis Abnormalities

I've mentioned situs already.

I refer to that as something being in the wrong place.

And this is a fetus with situs inversus.

This is an oblique cut.

So we have the chest up here.

We have the abdomen down here.

This is not a case of diaphragmatic hernia.

This is the stomach properly in the abdomen only.

It's on the right side of the abdomen and very close to the spine, which makes me think there's no spleen behind it.

The cardiac apex up here on the left side.

Here's another example where we have the heart, the apex is up here.

It was a very complicated abnormal heart patient didn't scan very well in the stomach located down here on the other side of the trunk, situs inversus.

And when you see that, think about major complicated forms of structural heart disease.

I mentioned the axis earlier.

The axis story is kind of an interesting one in that two groups reported the importance of the axis at the same time, and they were, the articles were published 90 pages apart from each other in Obstetrics and Gynecology in 1995.

One was Tom Shipp from Boston, the other was Ramada Smith from William Beaumont Hospital in Michigan.

They both reported exactly the same thing.

The normal axis of the heart is between 30 and 60 degrees from the midline.

So 45 is about the average, and it tends to be higher in abnormals, especially when there are outflow tract abnormalities.

So that's good because we don't see the outflow tracts, obviously, when we're looking just at a four chamber view.

And this can be a clue to looking at the outflow tracts.

It was also incidentally noted that the axis tends to be increased with both gastroschisis and omphalocele.

I don't think anybody has a good explanation for that, but it does appear to be a consistent finding.

Here are a couple of examples of abnormal axis.

In the left hand panel here, we've got a fetus with a pretty normal looking four chamber view, except if you draw an imaginary line bisecting the chest and think about where 45 degrees from that would be.

It's more in this kind of direction.

This to me, feels like it's closer to 90 degrees than 45 degrees.

And in fact, when we looked at the outflow, there was a VSD, the arrow was pointing actually to the aortic valve, it should say aortic valve.

This is the VSD just above it and to the right here.

And this was a very important diagnosis because it turned out that in addition to an overriding aorta, this fetus had trisomy 18.

And the clue that sent her in by her general OB/GYN to our office, was that four chamber view looking funny.

Here's another abnormal four chamber view.

Again, I think if you bisect the chest and think about where that ventricular septum is pointing, it's too far over to the left.

Otherwise, though, it's a nice looking four chamber view, and it's only when we look at the outflow tracts that we see the typical parallel appearance of the great vessels that we find with transposition of the great arteries.

And in fact, this is transposition with an intact interventricular septum, a critical diagnosis to make in a timely fashion, so that the fetus doesn't become deathly ill.

And some will succumb before a transport team can arrive to bring them to the central tertiary hospital.

It's very important to make this diagnosis prenatally.

We'll talk about transposition again in a few minutes.

Utility of Four-Chamber Screening

The utility of four chamber screening has been looked at in a number of different ways.

This was an interesting study a few years ago in Italy that looked at how good they were at identifying heart disease.

And they found about a third of heart disease overall.

They said about half of prenatally detectable heart disease, although they never really defined prenatally detectable, they did a cost analysis.

7,000 patients, total cost was $323,019.95.

Let me say that again.

There was $323,000 in 1995, and that worked out to $14,000 per positive diagnosis, which is a pretty good number.

It's comparable to the cost of identifying spina bifida or Down syndrome with our various screening studies.

If you take this one step further and think about 7,000 patients with 15% scan twice, that's about 8,000 scans total.

And if you divide 323,000 by 8,000 scans, you find that each scan cost roughly $40.

So the numbers in the United States today, certainly, and even in 1995, in terms of actual costs and what people were being reimbursed for doing basic obstetrical ultrasound, the costs would come out to be somewhat greater in the United States.

Toros, about 10 years ago, did a review of the literature looking at screening obstetrical ultrasounds and the value of fetal heart visualization and identifying fetal heart disease.

In Toro's paper, they had 8,300 patients of their own.

They found a number of other studies as shown on this slide that used screening to look at normal patients.

There were a total of 108,000 patients in those studies.

They left out the one third from the bottom here.

SHIELD didn't have a numerator or denominator that were good enough to calculate risk numbers.

But taking that one out, there were a total of 661 cases of congenital heart disease, which is a rate of 6.1 per thousand, and the overall detection rate, when you add all those things up, the overall detection rate was about 32%.

So their conclusion is about a third of cases of congenital heart disease were being screened.

These studies were from Europe, the Middle East, and the United States.

STO looked in detail at diagnoses of heart disease in different parts of Europe, found a range of detection.

And as you look at these, you can see that they diagnosed more of things that badly distort the four chamber view, which makes intuitive sense.

He also looked at the sensitivity and how it varied with the number of exams that were standard at different parts of Europe.

So in areas that had no routine ultrasound as their standard, only about 15 to 20% of cases were detected if one or two exams were routine.

About a third of cases were detected in places like Germany, where it is required by law to give women three ultrasounds in pregnancy.

55% were detected.

So that tells us a couple of things.

In a place where a lot of ultrasounds are being done, the people are probably better at doing obstetric ultrasound and detect more problems.

It also suggests that looking at several points in pregnancy may improve detection rate because things may evolve and look differently later in pregnancy than they do earlier in pregnancy.

Nuchal Translucency as a Screening Tool

I wanna take a couple minutes to talk about one other screening routine.

And that is the nuchal translucency, the use of measurement of the thickness of the neck to identify fetal problems.

And we use this as a technique for assigning risk of aneuploidy in the first trimester.

John Hyett, when he was working in London, suggested that you might be able to obtain very high detection rates at acceptable false positive rates for congenital heart disease using nuchal translucency.

In fact, these numbers from this particular slide from his early work in the mid nineties did not really pan out.

And what seems more accurate as this later data that he presented showing that the thicker the neck, the greater the risk of congenital heart disease.

And that's also been found in the United States in both the so-called FASTER trial and the BUN trial.

The detection rate of congenital heart disease was unfortunately low.

It was between 15 and 25% using nuchal translucency as the screening technique.

However, fetuses that had increasingly thick necks in both those studies had increasingly high rates of congenital heart disease.

So even though the relative risk goes up clearly and dramatically, it's not a good screening test for heart disease.

It's a good screening test for other things.

And if we're doing it, we shouldn't ignore the results.

But it can't be introduced as a primarily as a way of detecting congenital heart disease.

Benefits of Prenatal Diagnosis

The first description in the English language of using ultrasound to detect fetal heart activity and diagnose fetal heart problems was by Winberg in 1972.

Since then, we've gotten better.

We have better equipment, we have functional tests, but we, until recently, we didn't have an answer to the benefits of making what the benefits were of making a diagnosis of congenital heart disease.

I started thinking about this in the mid nineties after I gave a talk in Ontario and my title got messed up.

It was supposed to be fetal echocardiography makes a difference.

But somehow that isn't what made it to the posters that were all over the hospital.

And it got me thinking about how little data we actually had at the time to suggest real solid outcomes that there was any improvement for babies.

We then looked at our data at Yale and found that for at least for babies that were amenable to biventricular repair, there was a significant improvement in survival with a prenatal diagnosis.

And this paper published in 96 was the first paper about any form of fetal anomaly that showed an improvement in survival.

We looked for other things like length of stay and total cost, but did not find the expected improvements in those, in part due to some broadening of the standard errors due to a few complicated babies.

But at least in the study we did, we didn't have enough power to detect those differences that we expected other people have now confirmed tremendous improvements in outcome.

For example, Wayne Tworetzky, when he was in San Francisco, looked at their data for hypoplastic left heart and found that they got the babies to the operating room faster.

There were fewer babies with ventilatory problems, fewer that needed inotropic support and survival was substantially better in the group that had prenatal diagnosis than the group that were diagnosed after birth.

Bonnet looked at the results for transposition of the great arteries in France in 1999 and found again tremendous improvement in both complications and in survival for the babies that had a prenatal diagnosis.

And since these two studies, there have been a number of others that looked at variations on the theme of what morbidities could be evaluated and showed an improvement with prenatal diagnosis.

So it's become increasingly clear that we do better, at least in the short term.

And I think that the next frontier, and this is an important area for us to evaluate in future years, will be whether we can do better on the sorts of things that parents care about.

Beyond survival, how well will their child do?

Is there going to be will the child be normal or not?

And there are some data that have started to appear on the neurologic outcomes, looking at things like IQ scores for kids that have congenital heart disease.

You can see there's a pretty broad range here from the low nineties to the average at a hundred for babies with transposition, with tetralogy and ventricular septal defect.

The numbers look pretty good, probably normal for hypo for single ventricle, but for hypoplastic left heart, there is evidence for some kids of having really reduced full scale IQ scores.

And while that hasn't been uniform, this is an area that we need to understand better whether this is something that's prenatally determined or relates just to our postnatal management in the operating room, perioperatively, or can we have an impact by knowing about these kids before they're born and being able to optimize their status from immediately after birth until they go to the operating room?

Because many of these kids in the eighties and early nineties may have had to get very sick at community hospitals before they were transferred to tertiary hospitals to get their surgery.

And that could certainly have an impact as well.

So this is an area of investigation that I think really is important for the future.

Conclusion and Recommendations

So we've got a screening technique.

It's not perfect.

In fact, it's far from perfect.

And certainly, people should be thinking about adding the axis that's been out there for long enough that it should be part of our everyday thought.

You don't need to go buy a protractor.

You can look at it and get a sense of what the axis is like.

I think everybody should be trying to add the outflow views because it makes sense that the more we look at, the more we're able to see.

The current recommendations from the American College of OB/GYN, the American College of Radiology, and the AIUM are that whenever possible we should be looking at outflows.

There's some resistance to making that a mandatory part of the exam.

But I really encourage everybody to keep practicing and doing it all the time.

And if you're not sure, have somebody else look.

The more we look, the better we're gonna be.

NT is still to be determined.

I think it's going to turn out to be very helpful if it's being done anyway and it's thick, then fetal echo should be recommended.

We offer it in New Haven to women who have a nuchal translucency measurement of three millimeters or greater, although many people use 3.5 millimeters as well as a threshold.

Finally, I'd say that an imperfect screen is better than nothing at all.

If we don't look, we don't find anything.

Thanks.