Ultrasound of Diffuse Liver Disease - HD

Introduction

Hello, my name is Dr. Deborah Rubins. I'm from the University of Rochester in Rochester, New York, in the United States.

And I'm gonna be talking to you today about ultrasound for diffuse liver disease.

I have no disclosures.

Diffuse Liver Disease and Causes of Cirrhosis

When we think about diffuse liver disease, there are several things to consider.

One is how do we get cirrhosis?

And it can come from many causes including alcohol, viral lesions, non-alcoholic steatohepatitis, but carry syndrome and even hemochromatosis.

Some of these things are not necessarily identifiable by ultrasound, so we won't be talking about hemochromatosis.

Specifically today we'll be talking about the sequelae of cirrhosis by ultrasound, both portal hypertension and hepatocellular carcinoma.

And then we'll talk about some of the specific features of both steatosis and Bud Chiari that we can see on ultrasound.

Pathology of Cirrhosis

So cirrhosis is the final common pathway for many liver diseases and it consists of four things.

First, the hepatocytes are injured and they die, and then you get some fibrosis and inflammation around them.

There's attempted repair by the liver, which results in some nodules that are formed by the fibrous tracts.

And then we also get some volume redistribution, as the liver reorganizes.

So just to remind ourselves, the liver is organized around portal triads and you can see that very well in this graph.

Up here, the portal vein blood and the hepatic arterial blood come in at the portal triad accompanied by the bile duct and all the blood goes through the sinusoids and enters into the central vein.

So this is the pathologic correlate over here with the portal triad, the sinusoids and the central vein.

This is a larger depiction here, which is normal.

And you can see here portal triad, portal triad with a little bit of fibrous tissue, with the central vein in the center, between these rings of portal triads.

This is what happens when you get cirrhosis and you get this fibrosis that bridges between these portal triads.

The portal triad has collapsed and you can see that the sinusoids are basically obliterated and the hepatocytes are isolated within these bands of fibrous tissue.

Staging System for Cirrhosis

So there's a staging system that exists, for development of cirrhosis.

It was originally for hepatitis C, but now it's used in many forms of hepatitis.

And basically it's progressive fibrosis starting from the portal with just some inflammation and no fibrosis.

And then beginning with some fibrosis but no septa, which means no bridging in between the different portal triads, rare septa, numerous septa.

And then finally cirrhotic nodules, which is hepatocyte damage with bridging fibrosis.

Imaging Findings in Cirrhosis

The question for us is what can we see on imaging and with just portal inflammation just with lymphocytes in the portal trial as we really don't see much of anything.

But once we begin to get fibrosis, we actually can get some increased stiffness and edema in the liver and that can be detected by an elevated elastography modulus, which we can see with either ultrasound or MRI.

And you can see here as the fibrosis progresses, we can see this still with elastography, maybe with diffusion weighted imaging on MRI.

But until we get to cirrhosis, we really don't see anything on our standard gray scale ultrasound.

So just to show you the correlation here, here's an Mr ELAs gram, this is normal on the left abnormal on the right here you can see it's a cirrhotic liver with some nodule formation.

And notice here that when you vibrate the liver, you get these waves that are going through and you can see them in their different colors.

Here you get different wavelengths produced by the vibration depending on the stiffness of the liver.

And you can see here the wavelength is different as broader.

And when you calculate back from this wavelength into the speed of sound, the speed of the sheer waves that are produced, you actually can calculate back a relationship to the stiffness of the liver.

And so here's the sheer stiffness in kilopascals.

This is normal down at this low range here with an elevated stiffness, you have the patient who's cirrhotic and this has been plotted actually against pathology specimens.

And you can see here is normal down here.

And as the stiffness increases the VIR score, the fibrosis score increases as well so that we can clearly see a difference between cirrhotics and normal.

But we can also begin to tease out these areas of stiffness that we cannot see with standard ultrasound.

We can also do ShearWave imaging with elastography.

And this has been introduced, by many vendors over the past year with imaging.

So you can see you can generate some shear waves here in the right lobe of the liver.

You can put a region of interest on and calculate the speed of sound in this area.

And you take average of several measurements to get a mean kilopascal.

In this case it's 5.9, which is basically at the upper limits of normal.

But as I said, unfortunately except for elastography, we really have no way to detect pathology until we get to cirrhosis.

And so what we see on most imaging is the end stage of disease, which are the nodules, either small nodules under three millimeters or larger nodules that are greater than three millimeters.

And early on the liver will swell because of the inflammation you'll get hepatomegaly later on, the liver will scar and become small and fibrotic.

And in general, the right lobe scars first and the left lobe and the caudate lobe are spared.

So what do we see by ultrasound?

Well, we often describe this as a coarse echo texture, but what you really are seeing is the loss of the portal triads.

It's actually what we don't see that you're observing.

So when you look at this liver, we really don't see the portal triads that we should see out here.

And when you wanna see the nodules, actually you should go to higher resolution.

In this case, we went up to nine megahertz with a linear array transducer and you can see the nodules on the surface of the liver as well as some discrete nodules inside the liver.

Now Mr is much more exquisite at showing us these nodules.

You can see them here quite well because you can see the water in the fibrosis or the edema in the fibrosis sets it off against the background of the liver on T two weighted imaging.

And you can see very clearly the fibrosis separating this into multiple tiny nodules that may or may not be visible by ultrasound.

Confluent Hepatic Fibrosis

Now there's a special kind of fibrosis that it's important to recognize.

These are confluent hepatic fibrosis, which involves massive areas of the liver, in a fairly good proportion of patients.

And there are two typical locations.

One is the anterior segment of the right lobe and the other is the medial segment of the left lobe.

You can see the left lobe lesion here, which is low attenuation on ct, has some volume loss, of the capsule above it and does not enhance as you go through arterial and portal venous phase imaging.

Similarly on ultrasound, you can see that there's contraction here of the capsule and some, again, loss of all the portal triads in this patient.

But you see this area much better I think on CT here, where you can see not only the capsular contraction but also this sort of wedge-shaped area of decreased attenuation that becomes more conspicuous as you add intravenous contrast here in the portal phase, as well as in a delayed phase.

Sequelae of Cirrhosis: Portal Hypertension and Vascular Changes

So as I said, we're seeing the nodules of cirrhosis and the other thing that we see are the sequela of cirrhosis.

So because the portal triads have collapsed, there's no way for the blood that's coming into the liver to get out through the central vein.

So we lose the normal vasculature and we get creations of arteriovenous shunting.

We also get increased parenchymal pressure from that fibrosis.

And that leads to the development of portal hypertension, which you're seeing here with reverse flow in the portal.

Veins. Arteriovenous shunting is probably more of a problem for us on CT and MRI because we get enhancement in arterial phase imaging that you can see right here.

It's often peripheral, sometimes wed shaped, and it should enhance in the arterial phase, but then it should be iso attenuating on the portal phase, and should not wash out.

But this can be confused with early hepatic cellular carcinomas or dysplastic nodules and it can be very confusing, especially when it's round and has what we can see here as a draining vein.

Now this patient, the reason I'm mentioning AV fistulas is it's important for us to remember them by ultrasound as well.

This patient, came in with this lesion on MRI and he did have a cirrhotic liver and so biopsy was requested and when we looked at it with ultrasound, instead of seeing a discreet lesion, we saw this huge tangle of vessels, which you can see clearly has hepatic arterial waveforms as well as arterial waveforms in the portal vein that's draining.

So this is a very large arterial venous fistula.

And you can see here on angiography when you inject the hepatic artery, you get tremendous filling of this fistula as well as early draining into the portal vein.

So obviously you don't want a biopsy, this kind of lesion.

Now portal hypertension is much easier to detect, especially when you have flow reversal here as we see, so clearly on the color images as well as the spectral images.

A quick clue is the, opposing flow direction of the hepatic artery, which you see here.

And the portal vein, which we see here, you can see again the normal spectral waveform here in the hepatic artery.

And then the reverse flow in the portal vein.

So I have a couple of quiz cases for you, is this portal hypertension.

And you'll notice here that the left portal vein is going towards the transducer and the hepatic artery here is also towards the transducer behind it on the right side.

However, the portal venous flow is reversed while the hepatic artery flow here on the spectrum is antegrade.

So can this be portal hypertension with the left going forward and the right reversed?

And the answer here is when you look carefully, the flow is going towards the transducer, but it's not really going into the liver.

So when you look in sagittal view, you can see here is the flow in this vessel, it comes towards the transducer, but it's actually flowing away from the liver as well towards a very large umbilical collateral that you can see right here.

And here's the CT scan that supports that.

You can see, again, the flow in this vessel coming towards the anterior abdominal wall, but actually not going into the liver going into this CAPI Medusa of collaterals.

So portal hypertension does create collaterals because the blood has to get around the liver and back into the systemic circulation.

Most common is the coronary vein, but that's not specific for portal hypertension.

So other common collaterals include both lenal or intestinal veins in the retroperitoneum.

And a very common pathway is the patent paraumbilical vein, which is what we just saw, and that's seen in up to 20% of patients and is very specific for portal hypertension.

So that patent umbilical vein collateral will preserve forward flow in, in the sense of it is towards the transducer in the left portal vein, while the flow in the right portal vein is reversed.

Some other examples of collaterals here, you can see a patient with flow going towards the, towards the transducer here in the splenic vein and then away.

But when you look at the hilum here, you can see that there are multiple collaterals, in the splenic hilum and here they do drain into the left renal vein, which is clearly enlarged, and is communicating with systemic circulation.

We have some manmade collaterals, some that we don't see as often anymore, but you may recognize this.

This is a portal caval shunt.

And here we can see the splenic vein is flowing, from the spleen towards the IVC, but instead of going into the liver, it goes posteriorly back here and connects up into the inferior vena cava.

TIPS Shunts

The most common shunt that we use now is, trans esophageal int hepatic, portal systemic shunt, so-called tips.

And they are put in from the jugular vein into the right hepatic vein and then tunneled through the liver into the proximal right portal vein.

We look at these for their velocities as well as just their generic patency.

And you can see here is obviously a patent tips.

Here is one that's obviously occluded.

And the key to tips imaging is that all flow should be towards the tips IE it should always be towards the shunt.

So here is an unknown case for you.

Is this normal or abnormal flow?

And you can see here is a patient with the tips and when you look at the hepatic artery, it's going into the liver.

The right portal vein is reversed and the left portal vein is reversed going away from the transducer.

Left hepatic arteries going towards the transducer and the flow is going through the tips.

So is this normal or abnormal flow?

Well, the flow is reversed coming out of the liver, but we want all the flow to go towards the tips.

So this is right and left portal vein flow reversal in a normally functioning tips shunt.

So we monitor these patients post procedure to make sure that they're open and then, every three months and then as clinically indicated now with the newer covered tip stents, they don't have to be followed quite so closely.

These are the velocity ranges that we are looking for.

We'd like it between the velocity to range between 90 and 190 centimeters per second in the mid in distal shunt.

Remember, it takes a little time, for blood to accelerate and since it's coming into the portal vein at around 20 to 40 centimeters or second, it can be a little bit slower in the proximal shunt.

And so we don't use the proximal shunt flow velocities, as diagnostic for tips malfunction.

The other thing to look for is a, a change in stent velocity by over 50 centimeters per second over time.

Or else a doubling of flow within the tip shunt itself.

So just some examples here.

This is doppler signal from the, the portal vein here.

And then as you step up into the tips, you can see you get up to 70 centimeters per second.

It's not quite 90, but we're right at the beginning of the tip.

So this is certainly acceptable.

And as you move along through the tips, here's one at 70 and then it goes up to 1.2, 0.1 0.26 and then 1.08.

This is a normal tip with normal flow throughout.

On the contrary, this is an abnormal tips.

You can see the velocity is low in the proximal mid and distal shunt and the waveform in addition is very flat and non pulsatile, meaning that it probably is not communicating very well, with the right heart.

So this patient goes on to an angiogram and you can see before revision when you inject this tip shunt, there's a very, very tight stenosis here at the end.

But in addition, you can see all this filling of the liver, which indicates that the blood is not flowing towards the tips, it's actually flowing out into the liver.

After you revise the shunt, you can see that the stenotic area has been opened up nicely.

And again, when you inject all the blood flows through the tips and the hepatic blood that otherwise was being a pacified over here now is flowing towards the tips again and is no longer a pacified.

As I said, you can use elevated velocity, in the tips focally or certainly a doubling of velocity either will work.

And here this patient has a very elevated velocity in the distal tips of 256 centimeters per second.

And you can detect that by looking for the aliasing here, which tells you quickly where the highest velocity will be.

Now here's an area of of focal velocity elevation at the hepatic vein end.

You can see the velocity is 180, it's still within that 90 to one 90 range, but notice that in the mid tips it's only at 91.

So we've got a focal doubling of velocity and that should make us suspect that maybe there is some tips malfunction.

Notice that the portal vein still is flowing towards the tips.

So this is an early finding, and likely the patient is not in too much trouble yet.

But this patient went on to an angiogram and again, you can see when you inject that there's flow going into the liver.

And the patient had a gradient between the right atrium and the portal vein of nine.

When they angioplasty the tips and improved the flow, the gradient dropped to four and now there's again no filling of the portal vein branches.

So this is modification of an early lesion.

Now it's important to remember with the doppler that numbers aren't the be all and end all.

So here's a patient who has flow of a hundred at the hepatic vein end 68 in the mid tips and 57 proximally.

But notice this ity.

It actually looks exactly like the waveform you'd expect from the right heart.

So in the time when we did this study, we were acting on the basis of just the velocities and we sent the patient to angiography.

And of course the angiogram was normal.

So what I'll remind you to do is to also look for this pulsatility and remember that if you have a nice big stent, the velocity to get a certain volume of blood flow through that will be lower than if your stent is a little bit smaller.

So take everything with a little bit of a grain of salt and unless you have elevated velocities that are focal, or unless you have low velocities with flattened wave forms, be a little cautious with just using the numbers.

Portal Vein Thrombosis

The next thing that can happen when you have portal hypertension is you can get portal vein thrombosis.

And this is a frequent sequela because of the slow flow that's induced by the portal.

Hypertension, it may require color doppler to detect and as the portal vein clots you can have it recanalize or form collaterals and they should have typical portal spectral waveforms.

If you end up with tumor thrombus in the portal vein, that should have hepatic artery waveform and that can be biopsied for staging.

So let's look at some of these first an example of an koic portal vein thrombosis, you can see here are the portal veins.

Quite clearly they don't seem to have anything within them, but when you use color flow doppler at very low scale setting, we should see portal venous flow here and we're seeing nothing.

And here is the correlative CT where you can see all the collaterals here and you can see absolutely no flow in this portal vein.

Now acute portal vein thrombosis will actually distend the vessel a little bit, and that's similar to venous thrombosis anywhere you can see.

As this becomes more chronic, the main portal vein gets smaller and a collateral may form alongside of it as you can see here.

This also could represent the hepatic artery.

So you have to sample it, to be sure of which vessel is actually present.

Just another quick reminder you can see here, here's the hepatic artery with some aliasing 'cause we're sampling at very, very low scale here.

To pick up the portal vein flow, notice that the flow is going in bidirectional waveform here you've got flow going into the liver as well as flow that's coming out and that's not, atypical as patients slow flow develops.

One thing to notice here is that there's an area of disturbance here where the flow column is incomplete.

You can see it here on power doppler as well.

And that's actually a partial clot.

These are often seen much better with gray scale.

So don't be afraid to turn off the color because sometimes the color will bleed over these clots and disguise them from you.

Now everyone should have a portal vein and when you don't see one, the problem is it's probably gone.

So when you see too many of them, or too many holes that are black, you also have to turn on the color to figure out what's going on.

And here you can see multiple vessels in the portal system, which have portal venous wave forms.

And our diagnosis here should be cavernous transformation of the portal vein.

And this actually refers to these numerous collaterals at the porta due to either acute or longstanding thrombosis.

It can occur even as soon as a week after occlusion.

And if the portal vein is absent, you can consider using the superior mesenteric vein.

If you need something for a liver transplant, you just have to check that to be sure that it's patent.

Now this patient came in with a gastrointestinal bleed, and he came to us on ultrasound and the question is what's going on with his portal vein?

And you can see here that there's flow that's directed actually out of the liver and the sonographer looked at this.

You can see some more examples here.

You can see the flow going out of the liver.

This is the hepatic artery here for comparison.

So the question is what kind of portal vein clot is this?

And is this just portal hypertension with reverse flow and a partial clot?

Or is this something else?

And the key here is again in the waveform because you can see that comparing the portal vein waveform to the hepatic artery waveform, it doesn't look like a typical portal vein.

It's not flat and monophasic.

It actually has a regular puls ity which mirrors that of the hepatic artery.

And this is tumor thrombus of the portal vein.

And you can see here on CT here is the thrombus.

You can see the enhancement within it.

And when this was biopsied, this came back as hepatocellular carcinoma.

Now often the tumor itself is not very visible on either CT or on ultrasound and therefore, the biopsy is very useful.

Hepatocellular Carcinoma (HCC)

So speaking of paracellular carcinoma, it's a common cancer.

It causes a lot of year deaths per year worldwide and it's increasing in prevalence.

In western countries we diagnose 30 to 40% early stage.

And the five year survival for these patients is pretty good, 50 to 75% and we have lots of therapies to use to sort of locally ablate some of their tumors worldwide.

Most of the hepatocellular carcinoma is diagnosed in a late stage IE when it's already invaded the portal vein, or it has metastasized.

And the cure rate for advanced hepatocellular carcinoma is basically none.

So here's a patient with an elevated alpha-fetoprotein has hepatitis B, and you can see a very large mass here in the right lobe of the liver, not particularly vascular, but on the CT you can see it has arterial enhancement.

Notice that there are two other foci here in the left lobe as well.

And then on the more delayed portal venous imaging, there's areas that have decreased enhancement, both in the left lobe as well as in the right lobe.

And that's typical of hepatocellular carcinoma.

So how do we detect these earlier?

Well, the answer is screening and ultrasound is actually the vehicle for screening worldwide.

And screening is considered effective if the intervention that's performed will increase longevity by a hundred days at a cost of less than $50,000 a year.

So how does that relate to hepatocellular carcinoma?

Well, that actually occurs if you have an incidence over 1.5% a year.

So we're gonna screen people who have a high likelihood or high risk for developing hepatocellular carcinoma.

And screening is worthwhile again when we discover the HCC early on.

So the American association for this, for liver diseases recommends, surveillance for HCC using ultrasound because alpha-fetoprotein is not sensitive or specific enough and the A FP finds tumors at an increased stage when the treatments doesn't work very well.

So ultrasound all overall sensitivity not terrific, 65 to 80% but the specificity is high, which is useful because these patients get lots of nodules and other things going on in their livers, that you may not wanna detect.

And the other thing is that it can be difficult to know what to do with nodules that are a centimeter or less in size.

The recommendation is to screen patients at six month intervals.

And that's better, in terms of survival when compared to a 12 month interval.

And you don't have to shorten the surveillance interval for patients at higher risk of paracellular carcinoma.

So what do we see as HCC develops?

Well, it starts along a spectrum of trying to repair and as we try to repair the liver, it first forms degenerative nodules and then the nuclei become more, abnormal and atypical and you get a dysplastic nodule.

Then later on you'll get an early hepatocellular carcinoma and then it becomes a well differentiated HCC, moderately differentiated and poorly differentiated.

And these are pathologic terms.

What it looks like really for us is you have these areas that are dysplastic foci and if they have more hepatic arteries than they have, have portal veins, they are gonna be classified as an early hepatocellular carcinoma.

If they still have some portal triads along with their hepatic arteries, they're usually gonna be dysplastic nodules.

And once they've lost all their portal veins and they just have hepatic arteries, they become progressed to paracellular carcinomas.

Now you might say, what difference does that make on ultrasound to us?

We really can't see that.

Well, we can't see that with our standard gray scale, but we will be able to look for that when we are using ultrasound contrast, as many people do all over the world.

This is what we do see when we're using contrasted studies on either CT or mr.

And here is a well differentiated HCC pre contrast, slightly lower with comparison to the background of the liver on CT enhancement in the arterial phase because it has more hepatic arteries in it than the rest of the liver.

And it becomes hypo attenuating on the portal phase because it has less portal flow than the rest of the liver.

So we actually use the pathophysiology of the blood supply to tell us what's going on with our imaging.

And so these are the criteria for hepatocellular carcinoma.

Using CT or MRI increased late arterial phase contrast enhancement with a washout on the portal venous phase or peripheral rim enhancement or growth.

And that's by 50% or more on CT or MR that are less than six months apart.

And if your lesion is between one and two centimeters, you actually need all of these criteria.

Whereas if it's greater than two centimeters you just need the arterial phase contrast enhancement and one of these criteria.

What about on ultrasound?

Well, on ultrasound unfortunately the nodules can be a variable genicity and they may or may not show increased vascularity.

It just depends on how many arteries they have and what size they are.

So here's one that was seen on ultrasound detected by ultrasound and then monitored by ct.

And again, we can see the arterial enhancement here and also the growth six months later it's almost doubled in size.

Just another example here for you for variable appearance here is a lesion, in the right lobe.

Here's a closeup of it, it's hypo coic and it has some through transmission and it has some arterial waveforms within it.

So clearly it's an abnormal nodule, in a cirrhotic liver.

Here it is six weeks later and the difference is sort of striking.

It's quite iso coic here.

You can still see a little bit of that transmission and you cannot see any of the vascularity whatsoever.

So just difference in scale, slight differences in technique, can make a big difference in how the lesion looks.

And yet this is a typical lesion by MRI.

So what happens to these nodules that we detect by ultrasound?

Well, the practice guideline tells us if they're under a centimeter, just repeat the ultrasound.

If they're growing then you investigate them once they get to be bigger than a centimeter.

If they're stable, you just do surveillance ultrasounds.

Again, once they get to be bigger than a centimeter and you've found a nodule in a cirrhotic liver, then they go on to either CT or MRI.

In my practice we would probably use MRI.

And if they have typical features, then they can be listed for transplant.

If they don't have typical features, some of them will go on to biopsy.

There are some risk factors with this.

If the patient has a hepatocellular carcinoma, with a fairly good portal venous blood supply, it won't show that washout.

That will be a false negative and requires growth.

But sequential imaging is useful for these patients because it can improve the sensitivity and reduce the need for biopsy.

And ultimately we would like to not do biopsy because that, seeds the biopsy tract and also can cause hemorrhage.

So if we have classic features on either CT or MRI, there is no need for biopsy.

And if it's greater than two centimeters or less than five centimeters, these patients will be listed for liver transplant.

If it's bigger than that, or if the patient has already a diagnosis here, we will perform localized a ablation to try to get that lesion, under control and to keep the patient from developing complications from that lesion to get them to transplant.

And we're gonna be monitoring these patients every three months at this point by CT or MRI to look for recurrence.

Non-Alcoholic Steatohepatitis (NASH)

I wanna close with a couple of specific diseases that actually also lead to cirrhosis.

The first of these is non-alcoholic steato, hepatitis or nash.

It's a big national health problem at this point in time.

It begins as non-alcoholic fatty liver disease and the prevalence is estimated anywhere between 20 and 33%.

But there's been some screening studies of patients where the prevalence was as high as 46%.

And in my practice I see many, many patients with what I would term a fatty liver.

Fortunately, only a small percent of these patients will progress actually to steatohepatitis.

But a recent study again suggested that there's a 12% incidence of steatohepatitis overall and therefore about a third of these fatty liver disease patients do go on to steatohepatitis.

So once they get to steato hepatitis, what happens to them?

Well, well some percent actually go on to cirrhosis.

And if you do, you have a chance of developing hepatocellular carcinoma or portal hypertension, bleeding, et cetera, of up to 46, 40 to 60% over five to seven years.

Another small percent or about a third I should say go on to the progression of fibrosis.

But still with nash alone between a half and two thirds will be stable or actually get better.

So, NASH is not quite as severe a disease as hepatitis B and hepatitis C, but it still is significant because of the vast numbers of patients who have it.

So what's our motivation for imaging?

Well, one is to detect non-alcoholic fatty liver disease, in part because they're at risk for cirrhosis, but also because they're at risk for the metabolic syndrome, which causes diabetes and cardiovascular mortality.

And we can also find patients who do have concomitant liver disease.

Do we want to quantify the amount of fat to monitor therapy for early disease?

Well, unfortunately the amount is not predictive of NASH or cirrhosis, but perhaps finding the patients with fat would be useful.

And lastly, we can maybe monitor disease progression including cirrhosis and as complications.

So what do we see on imaging?

Well with ultrasound, we see increased echogenicity and that's thought to be due to increased acoustic interfaces from intracellular lipid vesicles.

We don't see the normal vessel walls when it's mild.

You can see here we've lost the portal triads, and the kidney is dark compared to the liver.

As it becomes more severe, we have more and more reflection of the sound back towards the transducer with more attenuation distally and we begin to lose the echoes in the posterior surface of the liver.

And you can see as it becomes severe, we've lost the ability to see the diaphragm posteriorly.

Some people have suggested that we can use an APA index to look for s steatosis.

That's basically what I just showed you.

If the liver genicity is greater than the kidney, then we call a fatty liver.

It's not accurate for very mild fatty infiltration or even moderate groups and it doesn't work in a patient who's got renal disease because then your relative standard is impaired.

The other thing is that ultrasound B scan doesn't directly measure the fat content.

It really actually measures the scattering.

So we're not truly measuring the fat, however, we can use this as a qualitative comparison.

And overall our sensitivity isn't too bad.

The specificity isn't too bad and we're actually about as good as CT or maybe even a little bit better, because CT also has some problems with other metals and things, altering the liver attenuation and basically the HEPA index performs well for patients with fat greater than 30%.

What about a quantitative HEPA index?

Can we actually measure the amplitude of the echoes?

And, this has had better success than just the qualitative look, but notice that you do get sort of a different look, and different echo amplitude depending on the transducer that you use.

So I think you do have to be careful if you're gonna do this to have a standard that's set on, on a single image imaging frequency.

Nevertheless, the sensitivity of this has been very good.

90%, specificity is 84%.

This is from, publication from 2012 and the same group, which is from the NER Clinic, did this again in 2015 where they actually took the images just from their PACS system rather than from the ultrasound machines and they were able to detect fat at a 5% threshold.

Is the amount of fat important?

Well, it's not related to the risk of steatohepatitis or nash and ultrasound is insensitive to lower amounts.

So it's unknown really if we need to quantify fat.

I will say that the pathologist divided into thirds mild is five to 33%, moderate 34 to 60% and severe is above 66%.

So in certainly in large chunks I think ultrasound can do pretty well.

There are a couple of other things about fat that we need to address.

One is that there are areas of focal sparing that can mimic masses.

And the clue to this is that you've got something that's got a geographic shape that has a flat edge.

There's no vessel distortion.

The vessels pass through this normally.

And there are some common sites, the medial segment of the left lobe or areas next to the porta, areas next to the gallbladder and areas next to the falciform ligament.

And this is the corresponding CT where the attenuation is slightly greater in this area that is not involved by fatty infiltration.

Focal fatty infiltration can also be nodular as you can see here, in this patient.

You can see it in March. It's slightly apparent here.

Again, normal vessels go through it here in November.

You can see that more of the right lobe is involved posteriorly.

Again, normal vessels pass right through it and when it gets very nodular and bizarre looking, it can be quite confusing.

Even on ct you'd have trouble deciding if this is fat or is this something else.

But by ultrasound there's no question this is fat with severe, severe attenuation.

And notice that this patient actually gets better quite quickly.

This is, a study in January by March the CT is completely normal again and the ultrasound is also normal when the patient is re-image, later in the year.

So fat facts, the infiltration can be diffuse or focal, it can come and go fairly quickly.

The typical spared areas are around the gallbladder and the porta.

MR can be used if you need to to differentiate focal fat from other hepatic lesions 'cause you can suppress the fat with MRI fat quantification is difficult with ultrasound and of unknown clinical value.

It can certainly be done with MR as well.

Steatohepatitis and nash and cirrhosis are a small fraction of the fatty livers, but as fatty liver disease increases, they will too.

Budd-Chiari Syndrome

And let's finish with one other cause of cirrhosis that we can detect by ultrasound and that's but Chiari syndrome, which fortunately is rare, but has a, a similar pathway to the development of liver disease and has some features that we can detect by ultrasound.

So here on ct, this is a typical image.

The idea here is that venous outflow is obstructed from the liver and the caudate has its own outflow passage.

So, it is t typically spared and the rest of the liver is involved.

So the pathophysiology is you have venous outflow obstruction either from the hepatic veins or from the IVC and therefore the pressure backs up in the sinusoids.

And that gives you portal hypertension, it gives you congestion around that, that central lobular vessel and the hepatic side die.

You can get ischemia and fatty change as well.

And lastly, you're gonna get central lobular fibrosis and the lobules will collapse and then the nodules will regenerate and you're into your cirrhosis pattern in Bud Chiari.

About a third are idiopathic.

And of those that we can find a cause, there are primary causes which are membranous webs that obstruct either the hepatic veins or the IVC and much more common are secondary causes including hypercoagulable states or inflammation or infection or even tumors.

Just some examples here you can see a clear web here crossing the IVC.

You can see the dilated IVC proximal to it.

And in the spectral tracing you can see a very flattened monophasic waveform.

You can see this on the cavo gram as well.

And notice on the CT that the IVC is very difficult to see in the hepatic veins are very attenuated.

Here's another patient again with ultrasound flow.

You can see the flow is actually reversed in the IVC going away from the web.

This patient had two previous dilatations.

You can see here there's absolutely no flow in the, this portion of the IVC on the MRI.

And when you inject this patient, you can see all these collaterals, getting the blood back through the systemic circulation, back to the heart.

Now when the hepatic veins are obstructed, it's usually due to thrombosis and if you see reverse hepatic vein flow, that's diagnostic, we also can cause hepatic vein obstruction from, for example, tip shunt, or even liver transplant.

Fortunately that's rare.

If we have a mechanical cause, such as the tips or liver transplant, stenosis, we can actually balloon angioplasty that or stent it.

So just a reminder, the normal hepatic veins are gonna have this nice right atrial wave form, two jots forward, one jot back.

Whereas when they're obstructed they tend to be damped like this and you lose the nice pulsatility reflected from the right heart.

On imaging, we see acute bud Chiari in about a quarter of the patients and that gives us a large liver abdominal pain.

In ascites the classic triad.

More often we're gonna see chronic changes in the patients who have already developed fibrosis and are forming peripheral atrophy and then they get regeneration with caudate enlargement and central hypertrophy and cirrhosis.

So the acute patient, as you would expect, has a very large liver and a large spleen.

And this one also has a trace of ascites.

And when we looked at the hepatic veins, you can see these very distorted disrupted hepatic wave forms that are not forward.

These are not phasic, these are just two and fro flow in the hepatic veins.

Conclusion

So in conclusion, cirrhosis is the final common pathway for liver insult.

There are multiple causes, viral alcohol, non-alcoholic steatohepatitis, but chiari, even hemochromatosis that we didn't discuss today.

Radiology monitors at this point for the sequela, for portal hypertension and for paracellular carcinoma.

And early monitoring is possible for, fat and for iron.

If you use MRI, if you're looking for hemochromatosis, but it's being investigated actually for reversible fibrosis in most of our cirrhosis patients by using elastography or again on MRI Elastography or diffusion weighted imaging.

Thank you very much.