The Use of Ultrasound for Screening for Down's Syndrome

Audio/Video Link  *requires RealPlayer - free download

Dr. Arthur Fougner: "Art Fougner here and with me is Dr. David Nyberg from the Swedish Hospital. Dr. Nyberg, I'm curious, how did you get started in the use of ultrasound for screening for Down's syndrome?"

Dr. David Nyberg: "First of all, I want to say it's nice to be here in sunny Orlando, and you've got a great shirt on by the way, Art. It really has to do with where I'm at. At Swedish Medical Center we are a referral center for the local area and the perinatal folks there do a lot of genetic amniocentesis and a lot of counseling. We see a lot of patients for referral ultrasound and the protocol has been to get an ultrasound for any patient going for amniocentesis or considering amniocentesis. So just naturally we have a lot of experience in evaluating patients so it just has evolved that way."

Dr. Arthur Fougner: "Do you think it's a reasonable thing to go from first studying things in people who somebody already thought to be high-risk to just kind of extend this into the so-called low-risk population?"

Dr. David Nyberg: "Yes, but that's kind of where the questions are now - how do you apply it to low-risk women? But that's been true, Art, for anything we've studied and not only in ultrasound but medicine and that is we've always started off with high-risk patients and then seen if we could apply them to low-risk patients. That seems to be the controversy now about how to apply what we've learned in high-risk women to low-risk women. I have no question in my mind we can do that, we can do it reasonably, and we can help screening performance but I also have no doubt that it could be used irrationally and, therefore, it can cause needless patient anxiety and maybe even potential harm. But I do believe that it can be used in that group."

Dr. Arthur Fougner: "It's been pointed out by some people that at least for some of the markers that perhaps the developments in the technology are responsible for noticing a lot more of these things. Have you found that to be true?"

Dr. David Nyberg: "So the question is did technology really improve our detection rates? To a minor extent it's true but I haven't seen a huge difference in technology and our detection rates over the last ten years. Maybe there have been some subtle differences but I do think that we pay attention to things more, I think more consistently the equipment is better, and I think our best equipment then isn't so terribly dissimilar to our equipment now. In fact, we probably still have some pieces of that equipment around. But I think more consistently that's true across the board if you look at all of the hospitals and communities and so forth, I'm sure that we've got better equipment than we did as a group then we did back then. At our particular center, I think we had good equipment then and, again, I'm not sure it's all that dissimilar."

Dr. Arthur Fougner: "Are there any signs that kind of leap off the screen that really are significant in being suspicious?"

Dr. David Nyberg: "Again, that's a good question. I think it's important to try to tell that difference. Let me just start back and say, first of all, it would be great if we didn't need sonographic markers and we could see structural anomalies - something very definitive in these fetuses particularly with trisomy 21. We can more or less with trisomies 18 and 13 but not with trisomy 21. Less than 20% of fetuses with trisomy 21 are going to show a structural defect with the exception of a few hands that's going to be true. But if we look at the combination of structural defects and markers, so-called genetic sonogram, we're seeing detection rates anywhere between 60%-80%. Now the false-positive rate is up there somewhere between 5% and maybe 15% but these are in high-risk women. So when we look at these various ultrasound markers, clearly some are better than others and we can look at the strength of those ultrasound markers by likelihood ratios. That's exactly equivalent, you can look at the strength of an ultrasound marker with likelihood ratios and that will at least tell us the power of that particular finding when we see it. It doesn't tell us the sensitivity, that is how often we do see it, but at least it tells us when we do see it how strong of a marker that is. So we've looked at isolated markers or looked at likelihood ratios for markers and it's quite interesting that for our isolated markers it's very similar to data that have just been published by the JAMA article. I have a lot of criticisms about that particular article but their data and their data summary is not bad. It's actually a very good piece of work in terms of what they've summarized. One of the things that they did was derive likelihood ratios for isolated markers, and we have a work in progress that will be published in the White Journal that those likelihood ratios are very similar to what we've seen at our own center, a single center. It takes a large number of patients before we can calculate likelihood ratios as an isolated marker so studies that look at 10-50 cases are not going to be able to talk about the significance of isolated markers. We had 180 patients; I think you're going to need at least 180 affected fetuses with Down syndrome. We had over 8,000 controls, and you need those kinds of numbers."

Dr. Arthur Fougner: "A very impressive database."

Dr. David Nyberg: "The meta-analysis by this article just out in JAMA by Smith-Bindman can look at an even larger number and, therefore, they can come up with likelihood ratios. Again, it's quite interesting that they're very similar but these smaller studies that try to tell whether it's significant or not simply can't address the question. We need big numbers."

Dr. Arthur Fougner: "Do you see a future for 3D? I just thought with the fetal hand open, perhaps, with the surface rendering to look for the simian crease."

Dr. David Nyberg: "That's interesting, we have picked markers to look at that can be seen during the course of a routine scan, very quickly that don't add a lot of time. We think those are the ones that are going to be most applicable - shortened femur, shortened humerus, and nuchal thickenings since we're going to look at the cisternal magna anyway. Obviously we're going to look at the heart so echogenic intracardiac focus, and we have several views that look at the abdomen so hyperechoic bowel makes sense. These things really don't take any additional time to speak of. There is a whole list of other things that we can look at but they're going to require more time. I think your idea of 3D is great, one of the things that's quite interesting with 3D is it's been clearly shown that a widened pelvic angle increases the risk but it's really hard with 2D to show that. We did a study with a 3D CT in post-mortem specimens and showed that was very useful and you could pick exactly your level. There is a paper that's coming out by Wesley Lee where he did the same thing with 3D ultrasound so there's another application of 3D ultrasound in picking another marker that could be useful but it's not something I would promote as a routine screen."

Dr. Arthur Fougner: "I think anything we can do to make a more definitive diagnosis and reduce both the false-negatives and the false-positives would be a plus."

Dr. David Nyberg: "I think that's great. If I can just take a second, let's address again this question that seems to haunt everybody about these single markers in low-risk patients because that seems to be the thing that's bothering everybody. First of all, at least with the markers that we use, we'll see single markers in about 11%-12% of normal fetuses. One of the ones that produce the highest false-positive rate is short femur so if we eliminated that we could decrease that. In fetuses with trisomy 21, the majority of fetuses will show more than one marker but in our study about 22% of all fetuses with trisomy 21 will show a single marker, so 22% versus 12%. In other words, even a single marker will increase your risk but you also have to look at the strength of that marker and, again, look at likelihood ratios. Let me just throw out an example because I think that and I'll use the data from Smith-Bindman. If you had a 25-year-old woman walk in your door and she showed an echogenic intracardiac focus - that turns out to be the single most common isolated marker in normal fetuses but also in our experience it's the single most common isolated marker in affected fetuses with Down syndrome. In other words, I don't think we can ignore it; on the other hand, we're not going to pay too much attention to it as an isolated finding. Using their data, they've got a likelihood ratio of 2.8 and a priority risk based on her age alone is about 1,040 for a 25-year-old woman. When you get done with the likelihood ratio for that finding, we end up with a risk of about 1 in 370 or something like that. I'd have to go through the numbers but about 1 in 370 and that would be a positive marker but a negative screen. That didn't cross anybody's threshold; it's very similar to biochemical markers. About 12%-15% of patients will have a higher risk from their age related risks because of the biochemical screen but until it crosses some threshold that we interpret, we'd say it's still a negative screen. It's the same thing so nobody should get excited about echogenic intracranial focus in a 25-year-old woman. Now on the other hand, I don't think they should ignore it because they should obviously do all those things anyway - look at the nuchal thickening and look at the heart. I can tell you not only from our antidotal experience but also summary of cases you're going to see echogenic intracardiac focus as the only finding in fetuses with trisomy 21. For those people that haven't seen that, it simply means they haven't seen enough patients yet or they're not doing their outcomes because they can tell you it absolutely happens. Not only that, we have seen patients with trisomy 13 where the echogenic intracardiac focus was either the only finding or the main finding that led us to the detection. So, again, it's not something you ignore, that can lead to a very valuable diagnosis, and here's something that increases her risk 2.8 fold. There are things in the headlines all the time where something may increase a risk 20%-50%; here we're increasing the risk 2-2.8 fold. If you look at Smith-Bindman's data, all of these markers increased from between 2.8 and something like 17 fold increased risk. How are we going to ignore these markers? What we have to do if we're going to use low-risk women is we have to apply it in a rational way."

Dr. Arthur Fougner: "Excellent, it's been a pleasure talking with you today."