Ultrasound Liver Elastography - HD

Introduction

Hi, I'm Richard Barr from Southwoods Imaging in Youngstown, Ohio.

Today I'd like to give you a lecture on ultrasound liver elastography.

The majority of this is taken from the SRU consensus document, which I think is a very good document, and a how to do a liver elastography, and I recommend that you review that document for additional information.

I do have several disclosures that are listed here.

The objective of this talk is to provide an overview of diffuse liver disease and discuss how liver stiffness values can be affected by populations etiology and other factors, and review the best practices for ultrasound liver elastography.

Chronic Liver Disease

Chronic liver disease is a substantial worldwide problem.

The inciting etiology of the chronic liver disease varies depending on which continents you're on.

So that in Europe, hepatitis C is the most prevalent.

In Asia we have both hepatitis B and hepatitis C.

And in the western world, non-alcoholic fat liver disease is the major disease that causes chronic liver disease.

The major consequence of all these diseases that cause chronic liver disease is increased deposition of fibrous tissue within the liver leading to the development of cirrhosis and its complications of portal hypertension, hepatic insufficiency, and hepatocellular carcinoma.

The stage of liver fibrosis is important to determine the prognosis, surveillance prioritization for treatment, and the potential for reversibility.

Now that new drugs are available for several of the factors that can cause cirrhosis and have been shown to reverse the degree of fibrosis, liver elastography is becoming extremely important to be able to evaluate these patients and to monitor treatment.

The Process of Fibrosis and Cirrhosis

The process of fibrosis is dynamic and regression of fibrosis is possible, again, with treatment of the underlying condition.

Previously, the only method of staging was of the degree of fibrosis, was by liver biopsy.

And serial liver biopsies are not something that patients would like to do.

So using elastography, in a large portion of the world, liver biopsies are becoming significantly less frequent.

Liver biopsy is really an imperfect histological reference.

It is been used as the standard gold standard for assessment of liver fibrosis.

It also allows us to stage the amount of information and degrade the degree of stenosis, necrosis and inflammatory activity.

And, unfortunately, a liver elastography is not able to distinguish all of these factors, but gives us an overall stiffness of the liver, which is a combination of fibrosis as well as inflammatory activity.

The liver biopsy is invasive and severe complications can occur in up to 1% of patients.

We really only biopsy about one 50,000 of the liver volume.

It is felt to be representative, but again, there are some issues if the biopsy can be taken in one spot that just happens to be a little less than and another in the degree of fibrosis.

And Ashley, there's considerable inter observer variability by pathologists, and studies have shown that the kappa value can range from anywhere from 0.4 to 0.9 in these studies.

This is a slide showing you what happens with fibrosis.

So, a normal or stage zero, we have our normal portal triads and hepatic veins.

And we then go through this area where we start to get fibrosis surrounding the portal triad.

And these then increase and then become bridging between adjacent sites and then more increased bridging, and then the development of nodules, and then a continued development of additional nodules.

There are three main scores that are used to grade these.

These are the ish aq, the var, and the bats.

Ludwig and I have these shown here.

The majority of studies use the ME score, which is a five point score with zero as normal.

And four as cirrhosis In a lot of studies, they like to use the is shag because it does separate mild cirrhosis from severe fibrosis.

So it has actually a seven point score, zero to grade six cirrhosis is considered the end stage of chronic liver disease of any etiology and all result from progressive fibrogenesis fibrogenesis.

So any chronic liver disease, and some of examples are hepatitis C, hepatitis B, alcoholic liver disease, non-alcoholic de hepatitis, cholestatic disease, as well as autoimmune diseases can cause this increased fibrosis.

And again, it's a continual process that leads to fibrosis and then to cirrhosis.

What Hepatologists Want to Know

What our hepatology colleagues are very interested in is knowing at which stage of this process these patients are in, because it makes them know exactly how to treat the patient, how often to follow the patient, and to give some idea of the patient's prognosis.

So, as you can see here, we have our chronic liver disease that leads to fibrosis.

We eventually get to compensated cirrhosis, and by compensated cirrhosis, I mean, the patient now has cirrhosis, me, stage four, but this is not easily diagnosed clinically.

These patients do not have very many presenting symptoms that the hepatologist will be able to do a clinical exam and determine that they have cirrhosis.

They have a median survival of greater than 12 years.

Also at this stage, they may or may not have varice and having varice tripled their mortality in one year.

So these are other in things that the hepatologist is very interested.

As this patient with compensated cirrhosis progresses, they have eventually develop varial hemorrhage, ascites, encephalopathy, or jaundice.

And at that point we call these decompensated cirrhosis patients.

This is an easy diagnosis, and usually the hepatologist can tell when the patient's walking in the room that they have decompensated and their mean survival is approximately two years.

So there's a huge difference between those patients that have compensated cirrhosis and decompensated cirrhosis in terms of morbidity and mortality.

And one of our goals using elastography is to try to find these patients in fibrosis or compensated cirrhosis so they can be triaged to treatment to prevent them from going to do decompensated cirrhosis.

Fibrosis also increases the portal hypertension so that using the metaverse score, you can see that we have normal pressures that have zero and they increase as we go to cirrhosis.

And again, one of the complicating features that is going to determine that the patient decompensates is going to be if they have bleeding viruses.

So the safest and most reliable method of measuring hepatic venous pressure is the hepatic venous pressure gradient.

And it's defined here, and it's usually three to five millimeters.

A hepatic venous gradient of greater than 10 millimeters per mercury is the strongest predictor of the development of varice patients with a hepatic venous pressure of less than 10 millimeters of mercury having 90% chance of not developing clinically decompensation in a four year period in patients with compensated cirrhosis, a reduction in the portal venous pressures of greater than 10% at one year protects against the development of varice and significantly improves outcome.

The non-invasive assessment of patients with chronic liver disease that our hepatology colleagues would like us to do, is there cirrhosis, is there not cirrhosis?

If there is compensated cirrhosis, does that patient have clinically significant portal hypertension or not?

If compensated cirrhosis is present or gastroesophageal is present or absent, and if they are treating this patient, is their improvement in therapy.

So when we design our report and our examinations, these are the things that our hepatologists would like us to address.

Spleen Stiffness

I do wanna mention that there is some early work on spleen stiffness as opposed to liver stiffness.

And the spleen stiffness appears to correlate with portal hypertension.

And here's a paper that again shows that the spleen stiffness actually correlates with increased hepatic venous pressures more than the liver stiffness does.

In another paper, they were able to actually predict which patients could decompensate by looking at the splenic stiffness.

And again, these studies are relatively early and have not been reproduced.

But I think in the future, we may also be doing spleen stiffness for those patients with advanced fibrosis or cirrhosis to again, help to characterize where on the spectrum of disease the patients lie to triage them to appropriate treatment.

Treatment and Regression

I just wanted to include these slides that just show that now that we have some medications, this zero for hepatitis C, that when these patients are treated with this therapy, they actually regress.

And the newer drugs coming out have shown a greater than 95% complete regression of disease and Hepatitis C the unfortunate problem is that these drugs are extremely expensive, and we need to triage who is the most appropriate to get these.

And at least in the United States, insurance companies have come up with one of the indications is a liver stiffness of a specific value to determine which patients should receive this therapy.

New drugs for hepatitis B as well as non-alcoholic fat liver disease are also coming out.

And again, so I think for the three major causes of cirrhosis, we will have treatment.

And again, I think these tests are gonna be extremely helpful in using liver stiffness to number one, decide which patients need treatment and to be able to follow treatment.

Importance of Staging in Disease Management

Why is staging important in the management of disease?

It's very prognostic.

So again, we can assign which patients are more needed in terms of urgency for the treatment and those that can wait given the cost, surveillance.

All these diseases lead to hepatocellular carcinoma and varice if they progress.

So we want to determine where they're at so we can again decide how the management will be given.

Again, we can use this for baseline monitoring of treatment deficiency, and we can tailor the treatment algorithms.

And I think this is going to be more and more important as we gain more experience in evaluating treatment of drugs with these applications to determine what is the most appropriate timing for follow-up examinations.

And again, this many of the insurance companies are now requiring that we have a liver stiffness value above a certain value before they will agree to pay for treatment.

Monitoring Treatment Frequency

How often should treatment be monitored?

Again, we don't know the answer to that because this is still in the learning phase, it may depend on what as the underlying disease.

So in other words, the response time may be different for patients with hepatitis C versus non-alcoholic fat liver disease.

It may vary with disease disease severity.

So patients that have an F three maybe need to monitor more or less than those patients that are starting with a metaverse F four.

It may depend on which type of therapy the patients are getting.

It may depend on what phase.

So is this in the induction of remission, or is this at maintenance?

After the patient has completed their treatment, we may be able to look at the toxicity and efficacy of treatment.

So it may be that instead of a normal 12 week course, some patients may be able to use an eight week course of drug saving a significant amount of cost.

And again, we may be able to use liver stiffness to help decide which of those patients can go to a shorter therapy.

And again, we can see the end points of where this is.

And again, some patients may be different than other patients depending on their phenotype.

So there we know that there are some patients that have a slow progression of disease where others have very rapid progression of disease.

So in general, if we had to guess at this point, elastography at one year intervals is probably not an inappropriate first guess as to what we want to do when the patient is being treated.

These may need to be a little bit shorter in time and patients on maintenance that have been cured, and maybe we need to do this at a longer time.

And again, I think research in the next coming years will help us decide and be able to give a better idea of what is the most appropriate for patient treatment.

Summarizing Report Content for Hepatologists

What I wanna do is summarize what the hep our hepatology colleagues actually would like us to include in the report.

So to them, they really put people in several boxes, and I think the three main boxes are the patient has no, or minimal fibrosis.

These are patients that probably need no immediate treatment, and the follow-up would be based on their initial disease and other clinical factors.

Those that have mild to moderate fibrosis.

In this case, the patients would be treated or follow up again based on other clinical factors.

So those patients that have a somewhere between an F two and an F three that may be appropriate for hepatitis C treatment, if that's their initial etiology, those patients would fall in this category and would be then triaged appropriately.

And then the third group are those that already have severe fibrosis or cirrhosis.

And this now triggers regardless of their initial etiology, of why they've got cirrhosis, that the patients need to be screened for hepatocellular carcinoma.

And in addition, to put them in the appropriate treatment arm, we need to decide if the patient has compensated or decompensated cirrhosis.

And in those patients that have compensated cirrhosis, do they have normal hepato venous pressures or are they abnormal?

And again, by placing patients in these different categories, it allows the hepatologist to determine how fast they need to treat the patients, what their prognosis is, and what their appropriate follow-up would be.

Liver Stiffness Calculations

So let's now move on to liver stiffness calculations.

Most of these techniques, if not all these techniques use sheer wave speed estimation.

So a sheer wave is just like other ultrasound waves as it goes through tissue.

Its speed is changed based on the stiffness of the tissue.

So the sound of the sheer wave will propagate faster in stiffer tissue and slower and softer tissue.

And to generate sheer waves in ultrasound, we use what's called an RFI, A RFI or push pulse, which is a long duration high intensity ultrasound push that's made to cause tissue to move.

Then we use tracking tissue displacement, our standard beam mode ultrasound perpendicular to that beam, where we monitor how the tissue moves.

And what you'll see is for a typical thing, this is the distance from the RFI pulse.

These are what happens to the tissue displacement.

And you can see closer to the, this is the distance from the RV pulse closer, we are to the RV pulse.

We get a quicker time to peak, and the peak is much higher.

And as we move away, the wave goes through at a later time, and we have some attenuation of the pulse.

If we take the peak time of displacement, and we plot that, we can then get the slope of this line corresponds to the ShearWave speed in meters per second.

This movie clip is a visual of what happens.

We're starting here and now where we got our RV pulse.

And you can see that this is the tissue displacement moves across the scan.

Elastography Techniques for Liver Fibrosis

Elastography for liver fibrosis.

There are several methods that have been used.

Strain is one method.

There's really a limited amount of things in the literature.

This has mostly been promoted by Hitachi, and they have very specific software that they use to calculate the liver stiffness.

Most of the guidelines do not recommend this be done at this time until further literature validates its use.

Then we have 1D ShearWave elastography, transient elastography by fiber scan, which is a one dimensional meaning we don't have an image, but the system monitors the ShearWave image.

Then we have point ShearWave elastography or PSWE.

In this situation, we use that RFI pulse, and we have a small ROI that we use to calculate the ShearWave speed in that small ROI.

We also have 2D ShearWave elastography, where we take many, many, many point ShearWave elastography in a very large field of view, usually color coding the results over that larger field of view.

And then you can take a smaller ROI and place it within that to get the shear wave speed at that point of the image.

There's also magnetic residency elastography, which uses a device that causes vibration.

And we look at the sheer waves as they go through the liver.

We using a specific m sequence, and we'll discuss these at a little bit more at length.

I think you have to remember the elastography measures stiffness.

It does not measure fibrosis only, and we'll talk about this at length as we go along here.

So that stiffness is influence both by fibrosis as well as increased hepatic pressures, including the portal hypertension, hepatic congestion, or increased blood flow from food digestion and also from inflammation.

So you need to be aware of that, and these are things you need to evaluate the patient for to determine if you're getting or cutoffs that are for fibrosis will be accurate.

Transient Elastography

If we look at transient elastography, it's been around for a long time.

There are a large number of studies and meta-analysis available, and these studies show that the area under the curve in indeed diagnosing cirrhosis is quite high a range of about 0.9 and classifies cirrhosis correctly in 85 to 94% of patients.

If we look at significant fibrosis, it does a little bit less accurate with an area under the curve of 0.75 2.93, with a correct classification of 57 to 90% in the there is substantial overlap between stages of hepatic fibrosis, particularly at lower stages.

And most of this work has been performed on hepatitis C, although more literature is now coming out for Hepatitis B and for nash, the advantages of using transient elastography, it's very easy to use.

The disadvantages are, it cannot be used if FAEs is present, and you cannot determine where the measurements are taken.

So you don't know if the patient has a large HCC, if you're actually taking the measurements of the HCC and not the underlying liver, or that you're also taking a measurement over a large vein, which will give you inaccurate measurements.

The probe that's used is a mechanical device that causes the thump.

And these need to be recalibrated every six to 12 months point.

Point ShearWave Elastography

ShearWave Elastography uses this phy pulse to generate shear waves in a small approximately one cc region of interest.

This is a realtime imaging technique, so you can use the realtime imaging to localize where you're taking the measurements so no masses or large vessels can be identified and avoided where we take the measurement.

You can systematically select different parts of the liver to do samples.

This is a newer technique than transient elastography, and there are fewer studies available, but the studies do show that this is a highly reproducible technique and meta analysis have found that point sheer wave shows a higher rate of reliable measurements and a similar predictive value to transient elastography for both significant fibrosis and cirrhosis.

This is just an example.

And in this case, again, we've got our standard B mode image of the liver, and then we can place our box in the liver and hit our button and we will get a readout as to what the liver stiffness is, and we'll talk about where we should be placing the box and how many measurements we should take in as we progress with this talk.

2D ShearWave Elastography

2D Sheray Elastography is basically doing multiple measurements using the RFI technology over a large field of view.

This can be done as a single image or performed in real time.

The real time imaging is again used on B mode, so we can determine where we're doing the lo the positioning of the box so we know where masses or large vessels are identified and void these so that we get more accurate measurement in general.

All these 2D techniques use a color coding that color codes the stiffness over the large field of view, and then allows you to place your region of interest.

It also allows you to vary that area of region of interest, so you can do averaging over a larger area of the liver.

Again, this is a newer technique and there's less studies out, but there are a couple studies that demonstrate that there is higher accuracy than transient elastography in assessing mild and intermediate stages of fibrosis using this technique and other studies that show that the 2D shear wave was more accurate than transient elastography in assessing significant fibrosis.

This is an example of 2D one where the liver stiffness is very soft.

And again, you can see here's our large field of view where we have measurements of the shear wave stiffness, and then they're color coded based on this color scale.

And in this image I've used ki past scales and this image I've used meters per second as the method that we're going to measure and we'll talk about which to use.

They actually are equivalent.

And again here we put our little dotted circle is our region of interest, and here is our readout of the mean maximum mean standard deviation of that value.

In this case, you can see we have an area that's red or very stiff liver, and much higher stiffness of the liver magnetic resonant imaging.

Magnetic Resonance Elastography

Again, a device is placed on the patient to generate she waves.

Special MR sequences is used to image these mechanical waves, and there's specific software that's used to color code, the image stiffness.

It actually is a relatively short exam taking about a minute for the acquisition.

And the results are displayed, is the sheer modulus.

And in all the ultrasound techniques, it's the young modulus that's displayed.

And the conversion from Young's modulus to sheer modulus is the sheer modulus is three times the Young's modulus.

And this MR technique allows for evaluation over a large volume of tissue.

So its advantages are that it evaluates a large area.

The disadvantages are obviously the cost of an MR is significantly more than that of ultrasound, and there's limited access to this technology.

Performing the Ultrasound Liver Elastography Examination

So how do we perform the examination?

An intercostal approach to the right lobe of the liver is preferred.

The patient should rise their right hand above their head.

To increase the intercostal space, you should use your beam mode imaging to find the best acoustical window.

That's where the beam mode signals are the best.

And there's no shadowing, because if you're getting the B mode ultrasound wave in, you're also going to be getting in the rv push pulse with less artifacts, and that's going to give you better signal to noise.

The measurement should be taken in a breath hold in a neutral breathing position.

This technique is very sensitive, and as you do a val Salva or take a breath in, you increase your hepatic venous pressure, which also increases the stiffness of the liver.

So we'll talk about this a little bit more later in the talk in more detail.

The measurement should be taken in the right lobe of the liver, as has been found that in many studies that the left lobe of the liver measurements are often unreliable.

And this is found to be that the motion from the heart significantly interferes with the ability to get an accurate measurement in the left lobe of the liver.

You also want to avoid the first one and a half to two centimeters from the liver capsule.

And the reason for this is there's reverberation of the aphy push pulse and the B mode pulses.

And again, we'll give you artificially elevated numbers.

So you should take your measurements approximately two centimeters deep to the liver capsule, not the skin, the liver capsule, to optimize that RFI displacement that's used to calculate the sheer wave speed and give us better signal to noise.

You wanna avoid large blood vessels and bile ducts, not only because their stiffness is different, but again, there may be reverberation of the sound waves from these vessels that cause an interference pattern and may give you inaccurate measurements.

You should also be concerned that it's not only vessels in that 2D view, but vessels that may be in or out of the plane.

So in addition to avoiding them in the plane scan a little bit above and below these to make sure that there's no large vessels near the volume that you're going to be evaluating.

And I believe that it's really important to get that aphy pulse perpendicular to the liver capsule, because if you have an angle, then you're going to get refraction of that aphy pulse, which again, decrease your signal to noise and provide you with less accurate measurements.

So here's just some examples.

What we want to do is, I prefer the patient obli.

Some people like to have them supine.

There's no difference.

I like the oly technique because I can rest my wrist on the patient because not only do I want the patient to hold still, I wanna make sure that I'm holding the probe very still.

You want to be parallel to the rib space so that, again, we're getting a B mode image without any artifacts or any rib shadowing.

So you want to avoid doing measurements at depth.

And the reason for that is that, again, the RRP pulse is attenuated, and the deeper you go, the less signal to noise you have, the less accurate your measurements are going to be.

In general at this time.

All of the vendors probably will not allow you to get a measurement at less than eight centimeters from the skin surface because there's just not enough RV pulse left at those points.

To get a reasonable measurement, you wanna avoid imaging add vessels as well as adjacent to vessels, because one, the vessel stiffness is different than the liver stiffness.

And two, you can get a refraction and reflection off the walls of the vessels that may cause interference with the measurement.

You want to avoid doing measurements at angles because again, that RV pulse may be refracted from the liver capsule and weaken it, and therefore give you less signal to noise and therefore less accurate measurements.

You don't wanna put your box right under the liver capsule because again, we know there are artifacts that occur from reverberation.

So you should try to put the box at least one and a half centimeters deep to the liver capsule.

When you do this on the techniques for the points your wave, it'll allow you to take many measurements and all the vendors have a package that you will save all the numbers, and then it will give you your average standard deviation and other measures that we'll talk about later.

And doing multiple measurements allows you to determine if your data sets are accurate.

Factors Affecting Liver Stiffness Measurements

There are many factors that affect the measurement of the rfe.

The amount of tissue displacement is dependent on the strength of the RV pulse, and we've discussed that a little bit already.

So that RV pulse is attenuated as it transverses through tissue.

Therefore, measurements taken at greater depth have less signal to noise.

And again, we said that I think all systems have a limit of eight centimeters, and I think the closer you are to the skin surface, but still maintain that one and a half centimeter below the liver capsule is probably the place where you're going to get the most accurate measurement.

This also means that those patients that have a very thick subcutaneous fat are going to give you less signal to noise.

They're gonna have more variability in their measurements, and also patients that have very stiff livers, the RV pulse is gonna be attenuated.

And again, in those cases, you're gonna have more variability in the measurements.

Breeding, I think, is one of the most critical factors that you have to control when you're doing these measurements.

So, as the patient breeds, the hepatic venous pressures change, as you know, from the hepatic waveforms when we do a doppler study.

So it's very important that the measurement should be taken in a neutral position that is no inspiration or expiration during a breath hold.

And I think this is really important.

'cause usually when you ask a patient to hold their breath, they actually take a breath in and then stop breathing.

And you don't want to have them do that.

You want to try to ask them to stop in mid inspiration expiration.

The measurements only take a few seconds.

It's very easy for most patients to be able to hold their breath.

So we like to actually practice this with the patients before we start taking measurements.

And again, try to get them so that they, when you say, hold your breath, they just stop breathing and don't take a breath in or out, and you ask them to do that kind of in mid inspiration or expiration.

The other thing that's very important is oftentimes when you tell patient to hold their breath, it may take them a second to actually do that.

So don't say, hold your breath and quickly push the button to do the rv 'cause they still may be moving.

Make sure you notice that the patient stops breathing before you hit the button to get your stiffness value.

So I showed you these two images before, this is actually myself.

And here I'm doing my measurements in neutral breathing where I've got a stiffness value of a 0.5 meters per second, which is normal.

And I did the best Val Salva I could do.

And there I've made the liver stiffness 4.8 meters per second.

So I actually went from normal to being cirrhotic just by doing a really good Val Salva.

And this is, again, just to show you the difference that breathing can make.

And being able to get your patients to always be consistent is going to be a big help in getting accurate measurements that are consistent.

We talked about variables, but there are several others that affect the sheer wave speed.

So acute and chronic conditions will also affect things.

So if the patient has acute hepatitis on top of their chronic hepatitis, the inflammatory process will also increase the stiffness of the liver.

So you're going to get higher readings.

And if you look in the literature, there are some correction factors, if you will, if the liver enzymes are abnormal that you can adjust the stiffness value to account for this.

If you have multiple disease etiologies, this confounds our measurements also postprandial is gonna give you higher reading.

So again, after you eat, you increase the blood flow to the liver and it actually increases the stiffness.

So we have all our patients fasting for four hours before we do the measurements, sampling on or near the vessels we talked about.

And again, remember in plane it out aplan because one of the problems is that there can be reflection of the sound wave from the vessel walls that cause an interference pattern and may change your stiffness value.

We talked several times already that we want to take the measurements at least 1.5 centimeters below the liver capsule.

And also don't take them too deep because the deeper you go, the less signal to noise you have, the more variability and measurements you'll have.

Sampling at the edge of a sector or sampling near rib shadows are going to also give you inaccurate measurements.

So you wanna avoid those.

And again, sampling in the left lobe of the liver, multiple studies have shown that this is much less reliable and it's probably because of motion from the heart.

We talked about motion, and again, this is patient motion as well as scanner motion.

So you really need to hold the probe very still and hold your hand very still during the acquisition time of these measurements.

And again, the patient, and again, you've got respiratory and cardiac motion that you really want to try to minimize and be reproducible when you do your measurements.

Other things like congestive heart failure or chronic renal failure also can affect the measurements because again, they have increased hepatic pressures because of the increased pressure in the right atrium from heart failure or fluid overload in renal disease.

So again, be aware of that.

So ideally this is what we like to do in our lab.

We want to have the RV pulse centered in the middle.

We want to have it perpendicular to the liver capsule.

We want to have a really good beam mode image and we want to place our rri I box about one and a half centimeters below the liver capsule.

Other things can affect the liver stiffness.

Extra hepatic cys stasis and use of beta blockers can actually affect the measurements period.

And again, Valsalva is a critical thing that can markedly change your measurements.

Units of Measurement

Should we be using meters per second or kilo pascals?

And the the measurements are exactly the same.

It's the same number meters per second is what we actually measure when we are doing these and all the techniques.

Kilo pascals is the physics term for this.

And to convert the meters per second to kilo pascals requires some assumptions.

For example, that the liver stiffness are the liver density is one, and that may or may not be true.

And presently all the systems use the same assumption, so it's reasonable.

Our concern is as this technology evolves, we may have additional parameters that may change the assumption.

So right now we prefer to use meters per second, which is what we're actually measuring.

And we think in the future when we're able to get the assumptions more accurate, doing kilopascals may be more appropriate.

Location and Number of Measurements

Should we take measurements in the same location or various locations.

Although liver fibrosis is a heterogeneous process, the best accuracy of liver ous values is from multiple measurements in the same location.

And again, that site should be the best location for the most accurate measurement.

One of the things I'd like to show you is all the vendors have this bias table, which is work of the key buck committee, which I'm not gonna discuss now.

But what this shows you at different stiffness values at different depths, there is a bias in each machine.

So if you decide to take measurements at different locations, not only are you adding the biases from the patient factors, but you're now entering also mechanical factors from the machine, which are gonna again, increase the variability and reliability of your measurements.

How many measurements should you take?

The literature suggests 10.

Should you delete bad numbers?

Some people say delete the highest and the lowest from a statistical standpoint, that's cheating.

And hopefully you would be getting really accurate measurements and not need to do that.

Some people suggest deleting measurements greater, less than one standard deviation from the mean.

And again, that's cheating and probably should not be done.

And if you're having to do that, that probably means you need to improve your technique.

I think as we standardize the technique, a smaller number of measurements should be appropriate and not out of line with the others.

Another measure is 60% of the measurements are good measurements.

And by that we mean if you get XXX or 0 0 0, depending on your machine, that the machine was not able to give you good ShearWave, if you didn't had measurements in six of those or 0, 0 0 or x, XX, then all the measurements are probably inaccurate.

And we see this in patients that have very, very stiff livers.

Because again, the waves the attenuation does not allow us really to get good measurements.

One thing that in the literature is I think very helpful for our lab is we use the IQR, the interquartile ratio divided by the median.

And that value of less than 0.3 suggests the dataset is good, meaning that the data is all within a very small range.

And we actually use this in the lab to monitor our sonographers quality.

And as we've been doing this over time, you can see that each stenographer's values have decreased as they've gained more and more experience and doing a better job.

And it also allows me to evaluate how we're doing in the lab.

And again, we've noticed a significant improvement in the values we get as we gain more and more experience.

And this is a really good measure for you to follow how your lab or even an individual tech is doing.

Determining Cutoffs and Reporting

One thing that I'd like to express is that a lot of the studies are trying to use the metaverse score to come up with cutoffs.

And we think that instead of trying to do that, using a likelihood ratio is actually a more appropriate, and by that, this is data from a large meta-analysis using point ShearWave.

And what we've done is taken their results and plotted them out the ShearWave speed versus the percent that was a specific metaverse score.

So you can see here's our F zeros, F ones, F twos, F threes, and F fours.

And you can see that particularly in the middle, there's a huge overlap between all these disease states or stiffness values.

And it's very hard for us to come up with cutoff values.

The SRU consensus panel had suggested thresholds.

And what we did is we went back to that initial table where I told you what does our hep hepatologist colleagues want to know?

So they want to know if there's no significant fibrosis or basically kind of normal patients.

So that we have a cutoff that's a low cutoff below that number, the patients are probably normal.

And then we have a high cutoff value above which everyone is most likely having either advanced fibrosis or cirrhosis and probably needs immediate attention.

And those in the middle, again, depends on other factors as to what they are going to do.

Also in the consensus panel, as we went through the literature and also gave cutoff values per each vendor.

So here I circled Phillips, you could go and these are the recommended values based on the literature if you want to use cutoffs based on the me our score, the SRU consensus best practices felt that you should report the median stiffness value and report some measure of quality.

And again, the IQR over median we think is the best measure and I think it would be very helpful.

Some vendors presently do not have this automated on the system, but they're all moving forward to have this value available on the system so that you have it and don't have to calculate it.

If you don't have it and you want to do it, you can just Google IQR and you'll get websites that have the equation or actually a little thing you can put in your measurements and it will give you these numbers for you to allow for improved reproducibility of serum measurements.

The patient position equipment, transducer, et cetera, should be tracked.

And that may not have to be in your report, but you should have it available.

So if the patient comes back, you are using the same equipment because there is variability between equipment and actually as we progress with in new software versions, we're getting better and better.

So you may want to maintain that information somewhere so you know if you've got changes secondary to equipment factors, and we really recommend that you discuss the doctor that you have, how you're reporting the results, so they know exactly how you're reporting 'em and how that fits into their practice.

So again, we chose to have two cutoff values.

One that is a liver stiffness value below for one vendor, 1.37 meters per second.

These have a low likelihood of fibrosis and are mostly f zeros and F ones and may not require additional follow up liver stiffness values of about 2.2.

Have a very high likelihood of cirrhosis.

And in that case, these patients probably should be seeing a hepatologist and probably need immediate treatment for their disease.

Then we have the group in the middle, which are kind of F twos and f threes, which are at risk for progression of the fibrosis depending on the etiology and other factors that the clinician would need to evaluate to determine how this patient should be treated.

Liver Stiffness and Portal Venous Pressures

At this point, liver stiffness.

Preliminary studies suggest that this may be very helpful in correlating with portal venous pressures.

So in those patients that are in that higher cutoff value, this may not be an inappropriate thing, but at this time we need more studies before we can recommend it for routine use.

Conclusion

So to conclude, detection of significant fibrosis and cirrhosis is important for diagnosing determination of treatment, prognosis, and follow up of cirrhotic liver disease.

The literature supports the noninvasive use of various EL elastography techniques to assess liver stiffness, to obtain accurate liver stiffness measurements, adherence to a very strict protocol is required and both patient factors and scanning factors affect results.

This is a slide that was from the SRU consensus statement that was put together by Dr. De Lavin from Boston.

And it actually outlines all of the issues that we have to deal with.

So we have pretest probability and post-test probability.

We got patient factors such as age, gender, ethnicity, and there are lab tests.

We've got a multiple number of diseases, all of which whose pathways lead to fibrosis and to cirrhosis.

We've got patient factors such as obesity, ascites, the medications they're on, and if they're fasting, we've got comorbidities.

We talked about acute and chronic disease or VA congestion, we've got different methods, Mr versus ultrasound.

And within ultrasound the variability of hardware and software technologists and interpreter experience and variability.

So all these affluence are elastography value.

And again, what we need to do is be aware of all of these and try to keep them constant.

So I'm gonna go back now to where we were initially when we said what does our hepatologist colleagues, what do they wanna know?

So now if we take that table that I showed you at the beginning of the talk, we can say no or minimal fibrosis.

So here we get the stiffness value.

And again, you have the table in this talk for the various vendors.

I've got one list here, then the mild to moderate, whereas the intervening values, and again, you can look at the table for individual vendors.

So these patients have mild to moderate fibrosis and need to have some treatment.

It may not be urgent.

But they do need to be followed up to determine if their fibrosis is progressing.

And again, those patients that may be in the category where we want to start treating patients with antivirals or treatment for their initial disease fall in this category.

If we move to the cirrhosis or fibrosis, those above the higher cutoff, now we have to start worrying that these patients are at risk for hepatocellular carcinoma and need to go into a screening protocol, which is usually recommended as a six month ultrasound.

We also want to try to determine if this patient is compensated or decompensated cirrhosis.

And we can do that by looking at other things other than our stiffness value that we do with our routine exams.

So those patients that are decompensated are gonna have ascites or bleeding varice, a rec cannulated umbilical vein.

Those that are compensated will not have those.

And we also want to try to determine if their portal venous pressure is higher or lower.

And again, a normal size spleen and normal portal vein are helpful to suggest that we've got normal portal venous hyper pressures.

If these are abnormal, then that's suggestive of abnormal hepatic venous pressures.

And again, the stiffness value in the future may be of some value.

So again, what we can do is use our stiffness values to place the patient in the appropriate box, if you will, of where they are in their disease progression, and then also use those boxes to determine what treatment and follow up is appropriate for them.

So I showed you this likelihood ratio and here's basically what we've done with our measurements.

So we've picked a cutoff value where we've got almost all of the normal F zeros and F ones.

We've picked a higher cutoff that has most of the cirrhotics and some f threes.

And then in the middle is this group that's significantly overlapping that have fibrosis, but not to the level of severe fibrosis or cirrhosis.

And with that, I hope this helps you to perform and interpret the results from liver elastography.

Thank you.