Ultrasound Liver Elastography: How I Do It - HD

Disclosures and Introduction to Chronic Liver Disease

I have multiple disclosures.

Chronic liver disease is a substantial worldwide problem. Its major consequences, increasing deposition of fibrous tissue within the liver, leading to the development of cirrhosis and its consequences of portal hypertension, hepatic insufficiency, and hepatocellular carcinoma.

Importance of Staging Liver Fibrosis

The staging of liver fibrosis is important to determine the prognosis, surveillance and prioritization for treatment. And now with the potential of reversibility with drugs, this is very important to decide when to start doing these treatments and to follow those treatments to see if they're working.

The process of fibrosis is dynamic and regression of fibrosis as possible with treatment of the underlying conditions.

Limitations of Liver Biopsy

And previously the only method of staging the degree of fibrosis was liver biopsy. And liver biopsy is really an imperfect histological reference standard. It does give us both a fibrosis assessment, but also con grades, steatosis necrosis, inflammatory activity, which we still cannot do with ultrasound, but hopefully that'll be coming in the near future.

The biopsy is invasive with severe complications and up to 1%, and we really only look at one 50,000 to the liver volume. So we don't get a very good sample. And if we look at the literature, a pathology that kappa values vary from 0.4 to 0.9 in studies. So there's very inconsistency on pathologists interpreting the liver biopsies.

Phases of Cirrhosis

This is just a table that describes cirrhosis. It consists of really two distinct clinical phases. First, we have fibrosis that leads to compensated cirrhosis, then to decompensated cirrhosis, and then death and decompensated cirrhosis. And by that I mean the patient has variceal hemorrhage, ascites, and encephalopathy and jaundice. So this is a really easy clinical diagnosis.

However, compensated cirrhosis is really not easily diagnosed 'cause it doesn't have any of those complicating features. And if you notice that the mean survival of a compensated cirrhotic patient is greater than 12 years, the median survival of someone with decompensated cirrhosis is about two years.

Role of Liver Elastography

So our goal in using liver elastography is to find these patients, either before they get to compensated cirrhosis with severe fibrosis or at the level of they have compensated cirrhosis so the patient can be treated. So they do not progress to decompensated cirrhosis.

In addition, varis play a role if there is no varis, the mortality and a, compensated cirrhosis is about 1% per year. If there's varis that increases by a factor of three to three C3 uh percent, and the elastography doesn't help there. But looking at our basic ultrasound image, is very helpful.

Methods for Assessing Liver Fibrosis

There are several methods of looking at liver fibrosis. Strain elastography is very limited in the literature, and I'm not gonna discuss that at all.

There are three ultrasound techniques. 1D ultrasound technique, which uses a mechanical push, or transient elastography. There's two RV techniques, point shear wave or 2D shear wave. So in point shear wave, we're looking at a small volume of tissue. And at 2D we're looking at a much larger area. And there's magnetic resonance elastography. And I'm not gonna discuss MRE today.

Understanding Elastography: Stiffness vs. Fibrosis

One of the key things I think you have to remember is elastography measure stiffness. It doesn't measure fibrosis. So the stiffness actually is influenced not only by fibrosis, but increased hepatic pressures. So anyone with portal hypertension, hepatic congestion, or increased blood flow from food digestion or even inflammation, will affect the values. So if we use the cutoff values, these things are going to affect where the cutoff value should be, or you may overestimate the degree of fibrosis of a patient has these situations.

Point Shear Wave Elastography

Points your wave. We use this phy or acoustical radial force impulse pulse, which is kind of a low high energy pulse that causes the generation of shear waves. Here we're looking at about a one millimeter area. It's a real time imaging, so we can look for masses and vessels and avoid those when we do our measurements. And it allows us to find a location or several locations to do these measurements.

2D Shear Wave Elastography

2D shear ShearWave is basically the same principle, but we're doing it many times over a larger area. Then we can put an ROI within that box, to get a reading. And the, the 2D ShearWave systems can either come as a single image, so you push the button and you get one image, or there's several vendors now that have this in real time so that we have a slow, slow refreshing, frame rate, of repeat measurements. And again, this is a real time technique so we can avoid masses and large vessels to find, identify the best place to put our box. And in 2D they all use a color coding assessment so you can get a general overview of the area. And again, this allows for, averaging every much larger area.

SRU Consensus Guidelines

This slide was made by Deb Levi from our SRU consensus panel. And basically in this one slide details everything we need to know. We've got all these different diseases that all lead to fibrosis and onto, cirrhosis. We have pretest probability and post-test, probability, age, gender, ethnicity, and lab tests can influence the values. We have patient factors such as obesity, ascites, medications, and fasting, comorbidities of acute on chronic disease and vascular congestion. And then we've got the several methods, either MR or ultrasound. And even between vendors, we get different values. So which software and which hardware you're using, as well as the experience of the reader as well as the performer of the examination. So to do this very well, we need to have a very strict protocol that's used all the time to get consistent results.

Performing the Examination

The way we perform this examination is an intercostal approach to the right lobe of the liver is preferred. The reason is if you do substernal or you try to look at the left lobe of the liver, you often apply extra pressure and you're also angling the probe, not perpendicular to the liver capsule, which causes some problems. And I like to say we wanna optimize the B mode image because that's telling us we got the best acoustical window. We wanna make sure that there's no shadowing. And if we're getting in the sound waves, we're gonna get a better stronger RV pulse. So we'll have stronger sheer waves, so it's easier to measure. And also we're always using the B mode image to track the ShearWave. So if you've got a bad B mode image, you're not gonna be able to track those shear waves.

The measurement should be taken during a breath hold in the neutral breathing position, and I'll show you a little bit more about that as we move through the talk. The measurement should be taken in the right lobe of the liver. I really don't tell my techs exactly where to do it. Usually it's going to be in segments five, seven or eight. And again, what we do is we try to find the best acoustical window that's going to give us our best results.

We also want to avoid the first one and a half to two centimeters from the liver capsule because we have refractive artifact that'll give us inaccurate measurements. So what we like to do is take the measurements about two centimeters deep to the liver capsule. And the reason is the farther you go down into the liver, the weaker you're going to get the RV pulse from attenuation. So you're gonna get weaker sheer waves, which is going to have it more difficult to get good measurements. The optimal place to do these is about four centimeters. So all the probes are kind of focused that your best is going to be at four centimeters. So if you have a person that has two centimeters of subcutaneous tissue, you've got the ideal situation because you always have to be about one and a half to two centimeters below the the liver capsule. So in those patients that have a lot of subcutaneous fat, it's very difficult to get good exams because you've got everything working against you. And we'll talk about what to do as we go through the talk.

You always wanna avoid large blood vessels and bile ducts. And one of the reasons is these have sharp borders and you get some refractive of the shear waves and you get an interference pattern, which will give you inaccurate measurements. And research that we've done that we really haven't published yet is you really wanna have the transducer perpendicular to the liver capsule in both planes. And the reason is you can really get a lot more energy into the liver by doing that. If you angle the probe, there's a lot of refraction of the energy, and then you're going to get weaker displacement and you're gonna have much more air in the measurements and it's very significant for small amounts of angling the probe. So again, you want to do the, you put your transducer parallel to the rib space and perpendicular to the liver capsule.

You wanna avoid imaging at depth. And again, the reason is the RV pulses, just like every other ultrasound beam, it's attenuated. The deeper you go, the more it's attenuated. The less the amplitude of the shear waves, the more air you're gonna have in those measurements.

You want to avoid imaging at vessels. And remember, this is both in plane and out of plane, so you wanna look up and down a little bit and make sure you're not near a vessel.

You wanna avoid imaging at angles. And again, the reason is the RV pulse is refracted, and you have less energy and you're gonna get less good measurements.

You wanna avoid imaging in the first one and a half centimeters to the liver capsule because of refractive artifact. And this is just an example showing that with this 2D ultrasound system, you can see that in this normal liver, which should be blue, you can see that we've got significant refractive artifact in this first one and a half centimeters. So this is one of the advantages. You're using these 2D systems. With the color maps, you can actually see how far that artifact extends. And depending on the patient, it's usually about a centimeter and a half, sometimes up to two centimeters.

Factors Affecting Measurements

There are multiple factors that affect the RV pulse and I look at to be able to get a good measurement, I always think about am I getting as enough energy in into the system? So the amount of tissue displacement is dependent on the strength of that RV pulse. It's attenuated, and therefore if we take measurements at greater depth, we're going to have more air in the measurements. So most systems will allow you to take measurements to from eight centimeters from the transducer, but we really try not to do that. We try to say again, it's about two centimeters below the liver capsule.

Another problem you're going to find is that the attenuation is greater with a stiffer liver or a steato liver. So there you're going to have, again, more issues. And it's not that steatosis at the way we're doing things now changes the measurement, but you just have more air in the measurement because you're just not getting enough energy into the system.

Breathing Technique

I think when you're starting out, breathing is the most important factor. It's very difficult sometimes to get patients to understand what you want to do. And you want to take the breathing in a neutral position. So if you look at your beam mode image, you'll see the liver moving up and down. You want to kind of take a position in the middle. And we really practice this with a patient before we start doing measurements. And the reason is, as you take a breath in, you actually increase the right heart pressure and it's transmitted to the liver and you actually increase the stiffness of the liver. So as you're taking breaths in and out, the liver stiffness is varying. So you what you want to do is make sure you try to do this pretty uniformly.

This is actually me and I did these images about 10 seconds apart. So you can see when I did things properly with my breathing, I had no fibrosis, but I became cirrhotic when I did the greatest of el cell I could do. So breathing makes a huge, huge difference. So you really need to practice this.

Variables and Vendor Differences

I'm not gonna go through the million variables that can affect the system because we're running outta time, but each vendor has their own frequency of the RV pulse, their own bandwidth of the RV pulse, the software that they're using to analyze the data, the technique they're using. So there is variability between machines. And at this point in time, you really can't use the same cutoff values for each vendor. So you need to go to each vendor and get the appropriate cutoff values.

Protocol for Measurements

For that, this is our protocol, again, one and a half to two centimeters below the liver capsule for this point shear wave with the 2D shear wave, we try to do this the same way also, but if you want, you can move the box up higher and you can see where the effect of the refractive artifact is.

Other things that affect the measurement are extra hepatic ccy stasis. The use of beta blockers is in the literature. I don't understand why, but it is, and again, as I showed you, vel, Salva can make a huge difference meters per second versus kilopascal every ultrasound system actually measures things in meters per second. And we make some assumptions for kilopascals. The FDA would prefer that we use meters per second. However, because FibroScan initially started with kilopascals, it's kind of been with the hepatologist and the drug companies that are paying for these drugs like to use.

So we usually end up in our report trying to give both. And I think most vendors, since the SRU consensus has come out, is now going to provide you both of those numbers.

When you do the measurements, should you take the measurements in the same location or various locations, we take 'em in the same location. So we do our 10 measurements in one location. And the reason is this is a bias table that we found from kiba that again, because the attenuation of the RV pulse and because there's a bandwidth, we actually get different measurements at different locations. So if you want measurements from different locations, you should take 10 measurements at each location. I dunno where 10 came from, but fibers scan started. It actually makes a lot of sense.

Do you delete bad numbers? I think one of the things that's really gonna help you is all the vendors are working on new quality measures. So you'll have maps that will tell you where is the best place within an image to take the measurement, or to where to ignore the measurement. And I think that's gonna really help you.

We use the IQ r over median of less than 0.3 to suggest this is good data. And I really love this measurement because I monitor our sonographers quality so I can tell you which sonographer is the best person to do this, which ones are struggling, and I really use that and I can tell that maybe when we started doing these routinely a couple years ago, we were more near 0.3, we're now down to about 0.1. So we've really improved and I can monitor that we're doing things. And again, we really noticed marked improvement in this value with experience.

Interpretation and Quality Measures

The one thing that I wanna tell you in interpretation, this is the likelihood ratio. So when we looked at a large meta-analysis is instead of trying to come up with a cutoff value, we plotted what was the probability of a certain metaverse score for a certain ShearWave speed. So if we take 1.3 meters per second in this study, you had a 2% chance of being of zero, a 10% chance of being F four, a 20% chance of being F three and a 35% chance of being F1 or F two. So it's really not realistic for us to think in the central area that we're going to do a good job in doing cutoffs, but we really do well with normals as well as abnormals. And that's why the SRU came with the cutoff values that we'll talk about shortly.

This is just an example of one of the ways of the quality map here. We're using a stoplight. Green is good, yellow is caution, red is bad. And here you can see the velocity map where you could do your measurements and here's this quality map. So what you could do is look at the areas that are green and say that's where to do your ROI.

SRU Consensus Cutoffs

The SRU consensus, like I said, came up with two cutoff values below which you have a high probability of being normal and a higher number, high probability of being abnormal. And we did that for each of the vendors that was available at that time. We also gave the cutoff values for the metaverse scores.

Best Practices: Reporting

So just quickly, to, since I'm running out of time, best practices, we report the median stiffness of the 10 values. This is kind of now what we're getting out from the vendors after the SRU that they're giving you both kilopascals and meters per second. And one of the things that we do that I have these macros, so for every vendor I have this little thing that we put in the back that shows you the SRU consensus cutoffs and the metaverse scores. And I think this is really helpful for the hepatologists, the refer us to us because they can convert from one to the other if they like.

And the other thing that I'm always ask is what do you do if there's a a, a problem? So after I give the number, and I usually just give the SRU, but things, if the patient comes in and they ate or they have got congestive heart failure, I'll usually add a sentence that says in the setting of patient eating, the degree of liver fibrosis may be estim overestimated. So I like to put that in as the things. If the measurements of the IQR ver median is higher, then we like to say something like the variance of the measurements is large and therefore the accuracy of measurements may be in question. So we add those into our reports.

And since I'm already over, I'm not gonna go through that. But here's basically what we did. We have this low cutoff below which everything is normal, a high cutoff above which you have a very high probability of having severe fibrosis or cirrhosis.

Conclusion

To conclude, detection of significant fibrosis and cirrhosis is important for diagnosing determination of treatment, prognosis, and follow up of chronic liver disease. The literature supports the noninvasive use of various elastography techniques to assess liver stiffness and to obtain accurate liver stiffness, measurements and adherence to a very strict protocol as required. And both patient factors and scanning factors affect results.

Thank you.