Screening for Hepatocellular Carcinoma (HCC) and an Introduction to LI-RADS - HD

Introduction

Hello, my name is Dr. Helena Gabriel. I'm from Northwestern University and the title of my talk today is Screening for Hepatocellular Carcinoma and an Introduction to the LI-RADS System.

LI-RADS, as many of you are familiar with, likely stands for Liver Imaging Reporting and Data System. It's one of the newer iterations of systems initially started by BI-RADS, which has been enormously successful in standardizing descriptions, reporting and tying it to management findings.

Thanks. This is to the success of RADS. We now have RADS, RADS, Lung RADS, RADS C RADS, which then begs the question, what's gonna be the next RADS.

This is the main RADS chart and from its inception it's been based on CT and ultrasound. There's been a problem, however because the main mode of screening for hepatoma is with ultrasound and there was no accounting for that. On this RADS chart, thanks to the newer RADS iteration, ultrasound now does have a role in this chart and before you get too excited, the role of ultrasound is here. It's a small role, but it's an important role and one that's likely to expand with newer versions.

Overview of the Talk

In my talk today, I'll discuss hepatocellular carcinoma, the problem evidence for screening for HCC, findings of HCC on ultrasound CT MRI. I'll then talk about LI-RADS and explain the system and show imaging examples and then I'll talk about how screening fits into the LI-RADS classification system.

Epidemiology of Hepatocellular Carcinoma

The hepatocellular carcinoma is not just a problem for select few, but it's worldwide epidemic if you will. The incidents vary strikingly by geography with global variations and incidents that reflect variations and risk factors, namely Hepatitis B and C.

This is a map from the World Health Organization showing the incidence of hepatocellular carcinoma in 2012, and you'll see that this overlaps greatly with the World Health Organization map showing the prevalence of hepatitis B highest in Asia, Africa, and the US as well as hepatitis C.

Hepatoma is the fifth most common cancer in men, seventh in women overall. It has the third most common cause of mortality worldwide.

This is a chart again, showing some of these global variations that we see in terms of the incidence of hepatoma highest in China, Southeast Asia, Europe, Africa, US. This also shows the incidence in mortality and hepatocellular carcinoma has a 1.07 mortality incident rate, meaning that if you contract hepatoma, you're likely to die from hepatoma if it's left untreated.

This is another chart that shows that the incidence of hepatocellular carcinoma continues to rise. And lastly, survival of hepatocellular carcinoma, especially in intermediate and advanced stages is dismal if left untreated.

Evidence for Screening

So with such a prevalent lethal cancer with surveillance and early detection help, and what's the evidence for this? There are two major liver societies, the American Association for Study of Liver Diseases and its European counterpart who are huge advocates for screening for hepatocellular carcinoma. And they actually provide an excellent evidence-based approach to screening diagnosis, staging and treatment in HCC.

In order for screening to be effective for any disease process, you first have to identify an at-risk population, and we know from the prior slides that hepatitis B and C are big risk factors for HCC. In addition, non-alcoholic fatty liver disease, alcoholic liver disease, aflatoxin, and any cause of cirrhosis independently is a risk factor for HCC. These are all patient populations that could benefit from surveillance, potentially.

Does screening for HCC decrease mortality? There have been a couple of trials, but the largest randomized control trial occurred in Shanghai where they looked at 18,816 patients who underwent six month AFP and ultrasound surveillance. And what they found was a 37% reduction in HCC related mortality. Even with the poor compliance that they experienced, only 60% of their patients adhered to the screening protocol.

Screening Modalities

Well, once you have a screening modality, how well does it do? What's the sensitivity, specificity and accuracy? In terms of screening for HCC, you can have serologic screening or you can have radiologic screening. In terms of serologic markers, AFP has been the most studied and that's what we'll discuss in terms of radiology screening. For HCC, it could be done with ultrasound, CT MR or ultrasound contrast potentially.

AFP as a Screening Tool

Let's talk a little bit about AFP. AFP is a glycoprotein secreted by fetal hepatocytes in HCC, usually poorly differentiated. The pitfalls of AFP though is that it can be high in chronic liver disease and hepatitis flares and not all HCC secrete AFP.

There have been multiple analyses looking at the perfect level and these have suggested that 20 is the best cutoff level, but even at this optimum level, the sensitivity of AFP for HCC is only 60%.

Because of this data, the new updated AASLD guideline states that AFP determination lacks adequate sensitivity and specificity for effective surveillance and therefore surveillance should be based on ultrasound exam.

Ultrasound for HCC Detection

Well, how good is ultrasound in terms of detecting HCC? And if you look at the literature, the sensitivity ranges anywhere from 30 to 100% with a specificity anywhere from 73 to 100%. There was a good meta-analysis that showed that there was a 94% sensitivity for detecting HCC at any stage and 63% for early stage HCCs.

The other finding that this study showed was that screening should be performed at six month intervals rather than one year intervals. Because median HCC double time is 170 days. So thanks to this data, the AASLD recommends a screening interval of six months for hepatocellular carcinoma.

CT and MRI for Screening

What about screening with CT/MRI? We know that CT/MRI have higher sensitivities for absolute lesion detection, compared to ultrasound, but there actually are very few studies evaluating CT/MRI in the context of screening for HCC. And there's no real screening data with MRI. Yet many institutions in the US do screen with CT/MRI.

There are some guidelines now that allow CT/MRI to be part of the surveillance program in patients in whom ultrasound is severely limited by obesity or other factors.

Cost-Effectiveness and Guidelines

Is screening cost effective? Of all the modalities in terms of imaging ultrasound is the most cost effective and it's probably this cost effectiveness and access that will likely keep ultrasound at the forefront of screening worldwide.

So thanks to all that information, the AASLD conclusion guidelines are that patients at high risk for developing HCC should be entered into a surveillance program. Surveillance for HCC should be performed with ultrasound MR and CT can be helpful in limited ultrasound screens. AFP is an inadequate screening test for HCC and is not recommended and that patients should be screened at six month intervals.

Ultrasound Technique for Screening

We're gonna go on now and talk about imaging findings of HCC Despite the fact that almost every major international society advocates the use of ultrasound for surveillance, there really are no clear cut guidelines as to how to perform ultrasound in patients with hepatitis and cirrhosis.

The ACR ultrasound guidelines give us a good start, but there's nothing specific regarding cirrhotic livers. The ultrasound screener really needs to be familiar with the appearance of cirrhosis and the cirrhotic architecture. And in fact, in some parts of the world there are specialized sonographers, hepa sonographers if you will, who are the only people that do these screens in cirrhotic livers.

In terms of our ultrasound technique and performing exams for hepatoma surveillance, we do numerous longitudinal and transverse static images through the liver confirming coverage of all aspects. We do longitudinal and transverse eclipse through the left lobe and the right lobe. Separately, we do high resolution surface images. We interrogate with doppler, the portal veins and we look at spleen length. We search for ascites as well as collateral vessels.

This is a typical exam of a patient with hepatitis B. We see that we have some very nice images of the left lobe of the liver and longitudinal. We then go into the right lobe. And in the right lobe we see that we've definitely gotten superior enough to see the diaphragm and lateral enough as well. So we have confirmed coverage of all areas.

This is another patient with hepatitis B and you can see that these sin eclipse are very helpful in showing coverage of all areas and making sure that we haven't missed any lesions.

Here's a different patient with hepatitis again, going through the left lobe, liver longitudinally transversely, we did the middle aspect of the liver here, laterally, superiorly, and again, getting liver surface images.

Not all livers are alike. The more cirrhotic a patient is, the more difficult it's gonna be and the more at risk they are for developing hepatocellular carcinoma. So it's very important to get coverage of all areas and to take a good look.

Here's a different patient who's very cirrhotic and these syn eclipse are very helpful in assuring us that we've gotten coverage of all areas of the liver and that we've done a pretty good screen here despite the fact that this patient has very coarse and architecture perception equals detection.

When you're looking for hepatocellular carcinoma, you should familiarize yourself and the sonographer should be familiar with cirrhotic liver texture and the appearance of hepatocellular carcinoma. Any technique that increases conspicuity in a disease, liver is helpful. B color can be very helpful. Scanning the liver in the nearfield and farfield separately can be helpful. Different dynamic ranges, harmonics, compound imaging and again, clips can be very helpful when surveying the liver.

This is an example of B color. We have a lesion over here, but it really tends to pop out a little bit with B color, making it more conspicuous.

Pitfalls and Missed Lesions in Ultrasound

Now ultrasound is not perfect. We saw that from the sensitivities that I had shown you on a previous slide. And there are lesions that can be missed on ultrasound. They can occur in areas of the liver that are difficult to image adequately, such as the lateral aspects in the dome of the liver.

The lesion may not be perceived in the background of cirrhosis. You can miss peripheral lesions due to the nodular lobulated margin of the liver due to cirrhosis. Isoechoic lesions can be very difficult to detect and infiltrative HCC can be difficult to pick up on any modality. Ultrasound, CT or MRI.

This is an example of a lesion that almost was not picked up. These were the initial ultrasound images here. What we saw was that we didn't feel like we were able to see this portion of the liver well. So we went back in there and we really tried to interrogate that lesion and indeed there was something truly there that you can perceive here and here the B color again pops it out a bit and the MR confirmed that indeed there was a lesion here. It's hyper enhancing. There's washout, there's restricted diffusion and it correlates with this lesion that we picked up on ultrasound.

Here's a different patient images of the right lobe. Perhaps there's a small hypoechoic lesion here, but this was not detected initially. And on mr, we see a hyper enhancing lesion with washout, which was a hepatoma, a different patient, a very lobulated margin to the lateral segment of the left lobe of the liver. Perhaps there's a hypoechoic area and indeed corresponding to this area is a fairly large lesion that's hyper enhancing with washout compatible with HCC.

Here's a different lesion. We see an area of heterogeneity in the right lobe of the liver, little mass like here, some abnormal color doppler imaging. The left lobe though looks completely unremarkable. This patient had an MRI because of the abnormal architecture here in the right lobe. And on the mr, we can see that there's a large infiltrating mass in the right lobe, but there's also a large infiltrating mass in the left lobe that really could not be detected even in retrospect on the ultrasound areas of high b800 signal as well.

I wanna just touch upon a pitfall that you can see if you are doing enough of these exams for a surveillance in cirrhotic livers. If you're doing high resolution surface images to look at nodularity and look at the texture. Inevitably, you may come across a situation where you see a small echogenic lesion. Here we have numerous small echogenic lesions, within the surface of the liver. And if you go to a lower frequency image, you won't see them at all. But they're very apparent on these high frequency images.

Now I'll tell you, when I first started seeing these, I did MRS on all these patients and the MR is inevitably negative. What do these lesions represent? We're not really sure. They could be hyperplastic nodules, they could be cirrhotic nodules, it could be biliary hamartomas. No one has really done path correlation on these lesions. Oftentimes they're multiple. Again, there's no MR correlate.

What we do is if we see these lesions, they're typically less than one centimeter. We follow the AASLD guidelines and we do a follow-up ultrasound in three months. If the lesions are stable, we then return the patient to routine surveillance, which is six month follow up. If they're greater than one centimeter, we then go directly to CT/MR evaluation.

So when you're done doing your ultrasound for HCC screening, you should ask yourself, was this an adequate screen for a hepatocellular carcinoma? And if not, perhaps you need to go to CT/MR and if it was an adequate screen, is there sonographic evidence for HCC?

Imaging Findings of HCC on Ultrasound

Let's talk now a little bit about the imaging findings of hepatocellular carcinoma and ultrasound lesions can be single, multiple or infiltrating the echogenicity can vary. A good portion of these about 50% will be hypoechoic, but they can be isoechoic. Hypoechoic or can be a mixture. They can be well-defined or poorly defined lesions. Sometimes there's a hypoechoic halo which may represent a fibrous capsule and these lesions do tend to be hypervascular.

What I'm gonna show you now are a lot of different examples of hepatocellular carcinoma. I will show you the ultrasound image. Most were found on ultrasound screening. Most of these patients had MRI correlation and many of these patients have path results thanks to resection or explanation for liver transplant.

Here's the first patient. This is a bit of a difficult lesion to see, but it is a hypoechoic lesion in the right lobe of the liver ill defined in terms of its margins. And this did turn out to be a 3.2 centimeter, moderately differentiated HCC.

Here's a different patient with a very small eight millimeter hypoechoic lesion right by the gallbladder. We see the MR correlate over here, a very small hyper enhancing lesion with washout. This was an eight millimeter grade one well differentiated HCC different patient with a hypoechoic nodule that actually bulges the margin a bit. It's a bit hypervascular as well an MRI. It correlates to this lesion hyper enhancing washout. This lesion was treated with radioembolization with a decrease in size and at explanation represented a 1.8 centimeter HCC necrotic, another hypoechoic HCC right by the middle hepatic vein.

This person also had an echogenic lesion in the right lobe of the liver On mr. This patient had a left lateral segmentectomy showing a 1.5 centimeter HCC. No lesion was found in the right lobe on mr, but the patient then went on to have a liver transplant and they did find a small right lobe HCC that was not detected on MRIA.

Different hypoechoic HCC as we see here. And here's the corresponding MR with hyper enhancement, a small tiny little HCC in the right lobe. And this is an isoechoic HCC. These can be very difficult to detect. Here's the hepatic parenchyma. Here's the lesion. Completely equal in echogenicity. No real vascularity. Perhaps a little bit in the periphery. This is the correlate on mr. Not very bright on mr, but definitely hyper enhancing with washout.

Here's a different isoechoic HCC, very difficult to detect, but here there's a definite contour bulge, which is a big clue. Here it is again peripherally hypervascular and on the mr. It corresponds to this lesion at the periphery compatible with an HCC. This lesion underwent radioembolization.

HCCs can sometimes have a hypoechoic halo which would suggest a clinically significant disease. And here's the lesion on a mar mildly hyperintense. T two hyper enhancing washout and with a capsular appearance. This lesion underwent transarterial chemoembolization and at explantation was a 2.2 centimeter grade two HCC.

Here's a different patient with another halo and a contour bulge. This was a very vascular lesion. Here's the corresponding mr. This patient also underwent TACE and radioembolization also had an explant showing a 2.5 centimeter HCC.

Some HCCs can be echogenic as well. We see this lesion here. A little bit of vascularity. Here's the corresponding mr. This patient underwent radioembolization.

Here's an echogenic HCC with a hypoechoic halo and a heterogeneously echogenic mixed HCC. Here this patient underwent radioembolization and had multifocal lesions. Rarely these lesions can be partially calcified as we see here. Here's the corresponding MR lesion.

So the typical ultrasound appearance of HCC is that they can look like anything. Many of them will be hypoechoic, but hepatoma can have almost any appearance. And again, perception is very important in the detection of these lesions.

Ultrasound Contrast

I'd like to talk just a little bit about ultrasound contrast. The AASLD put out a statement stating that contrast enhanced ultrasound may offer false positive HCC diagnosis in patients with cholangiocarcinoma. And as a result has been dropped from the diagnostic techniques.

Now, after the statement was released, there was a backlash of literature in Europe and there is now a LI-RADS ultrasound contrast working group that is looking specifically at this topic and trying to come up with new recommendations regarding contrast ultrasound as well as cholangiocarcinoma and its differentiation from HCC On ultrasound.

HCC Findings on CT and MRI

I'd like to talk a little bit about the findings of HCC on CT and MR and contrast ultrasound. And many of the findings that we see reflect some of the changes that can be seen with hepatocarcinogenesis of hepatic nodules.

Hepatocarcinogenesis is a transformation of normal hepatocytes to malignant nodules, namely HCC. This is the typical schematic that's often shown demonstrating changes of hepatocarcinogenesis from a cirrhotic nodule to a dysplastic nodule to HCC. HCC occurs as part of a hepatic field change. Cirrhosis causing fibrosis and nodular regeneration. Hepatocarcinogenesis is a sequence of de-differentiation from regenerative to cirrhotic to low grade dysplastic, high grade dysplastic early HCC to progressed HCC.

Now it may not be that all hepatoma evolve from the sequence, but it's thought that many do undergo these changes on their way to malignancy. And it also may be that these are not discrete steps, but part of a continuum.

So with this transformation also comes other changes that can have an effect on what we see on CT/MR. As there's more de-differentiation, there's actually an increase in unpaired hepatic arteries as we can see here in red. In addition, there's increased sinusoidal capitalization and there's a decrease in portal vein triads as well.

And here are some images showing some of these changes and what they look like on mr. At least we see arterial hyper enhancement. And again, this is probably because of an increase in unpaired hepatic arteries and sinusoidal capitalization that can occur with HCC. And the later on we see a washout appearance and that's because the drainage patterns have changed due to the fact that there are sinusoids perhaps causing early drainage. And there's no further portal vein enhancement due to decrease in portal triads. However, the adjacent parenchyma has normal triads so that there's a perception of washout of the hepatoma relative to the adjacent parenchyma.

Introduction to the LI-RADS System

I'd like to move on now and talk about the LI-RADS system and why do we need LI-RADS? Well, we need LI-RADS because it's a virtual alphabet soup of guidelines, societies CT and staging systems in liver disease. And all of these societies and systems had different nomenclature recommendations and descriptions.

LI-RADS tries to provide congruence and standardization with the criteria of all these systems, namely the AASLD and the UNOS OPTN systems. What LI-RADS is, is a system of standardized technique, terminology and criteria for interpreting and reporting findings of CT/MR exams in patients at risk for developing HCC, it was developed by an ACR appointed committee of radiologists with input from hepatobiliary surgeons, hepatologists, hepatopathologists, and interventional radiologists.

In other systems, imaging features were not precisely defined. In fact, in the first UNOS guideline for HCC HCC was defined as some type of vascular blush, extremely vague way of defining it. So there was ambiguity in the diagnosis of HCC. Additionally, prior systems resulted in a large number of indeterminate lesions and biopsies.

LI-RADS is more precise, allowing diagnosis of more benign entities and a more specific diagnosis of HCC. It also encourages follow up of lesions that are likely benign and all the while it aims for very high specificity. For HCC, there are five levels of suspicion for a nodule when using LI-RADS.

There is LR-1, which is a definitely benign lesion, 100% certainty of being benign and management should be just routine surveillance. There is LR-2, which is a lesion that's probably benign, has features that are suggestive but not diagnostic of a benign entity. Routine surveillance is usually done here.

LR-3 is an intermediate probability lesion for HCC. Features do not meet criteria for HCC and oftentimes these patients get followed. LR-4 is a lesion that's a probable HCC features are suggestive of, but not diagnostic of HCC. These patients can undergo close follow up additional imaging or biopsy.

An LR-5 is a definite HCC. According in LI-RADS criteria, there is 100% certainty of HCC treatment could be performed without biopsy and this lesion can contribute to T staging.

So again, LI-RADS strives for very high specificity for HCC and you can treat an LR-5 lesion without biopsy as a malignancy and management and staging can be taken as hepatocellular carcinoma.

Cirrhosis results in many nodules, malignant non-malignant. So that biopsy of these nodules would not be feasible. And in addition, there is a risk to biopsy. There can be tumor seeding after biopsy in about 2.7% of lesions. For example.

This is a patient who has a relatively large lesion in the lateral segment of the left lobe. It's mildly hyperintense T two. It's hypo enhancing on the early phases and remains hypo enhancing later on. Now because this never showed arterial enhancement, it was very difficult to call this as a definite hepatocellular carcinoma, though was quite suspicious because it was a large lesion. This lesion underwent ultrasound guided needle biopsy. Here's a needle tip within the lesion. Findings were that of hepatocellular carcinoma. The patient underwent radioembolization. We can see that the lesion has shrunk and is non enhancing, but we can also see that there are enhancing nodules within the interabdominal wall.

So the interabdominal wall was seeded by tumor at the time of the biopsy. This patient did fine because they did radioembolization of this tumor seeding within the interabdominal wall and obliterated these lesions.

LI-RADS Classification System

Here's a classification system and it appears a bit daunting, but I will break it up category by category and explain it. The category, the classification system first starts with an observation and we call it an observation instead of a lesion because not everything we see is an absolute lesion.

We then go to lesions that are treated, LI-RADS treated. These are patients who have had some type of liver directed therapy usually. So those are used as a different category.

LI-RADS also looks at lesions that are malignant, but non HCC malignant malignancies such as cholangiocarcinoma. It also talks about lesions that have tumor thrombus, which have important clinical and management ramifications. And LI-RADS is able to take these into a separate category.

So for example, this is a patient with a lesion in the right lobe of the liver. On the T two weighted images, we see some hyperintensity within the posterior branch of the right portal vein. We see that there's some low signal intensity here, but there is some probable enhancement on the diffusion images. There is increased signal. This is tumor thrombus within the right portal vein from the small lesion in the right lobe of the liver.

LR-1: Definitely Benign Lesions

The next category that LI-RADS will look at is those lesions that are benign. LR-1, there's an old adage that says good things don't happen in bad livers and that's not absolutely true. It's true that as cirrhosis ensues, there's a lower incidence of some of these benign entities because the fibrosis tends to obliterate them. But you can still have benign lesions in a cirrhotic liver.

And those benign lesions include cyst, hemangioma, vascular perfusion alteration, focal fat, pseudomass confluent fibrosis and scarring, as well as any observation that spontaneously disappears.

Here's an example of one such benign lesion. This is a well circumscribed lesion in the liver. It appears to be changing appearances appears echogenic here and more hypoechoic care. This lesion underwent MRI and on the T two weighted images, it's very bright. There is peripheral nodular enhancement on early enhancement sequences and then it entirely fills in. So this is absolutely compatible with a hepatic hemangioma, a completely benign LR-1 lesion.

LR-2: Probably Benign Lesions

Next we'll tackle the LR-2 lesions. Those lesions that are probably benign and these lesions include all entities that are benign but slightly atypical. These include the LR-2 cirrhosis associated nodules that appear somewhat distinctive.

So for example, here's a lesion that's hyperintense T two but does not demonstrate typical peripheral nodular enhancement, although there is some fill in this represented a sclerosed hyalinized hemangioma, which are difficult lesions to diagnose. If you don't have an antecedent MR or some type of imaging proving that it looked like a hemangioma before, this might be a lesion that we would follow.

Here's a different lesion, an LR-2 lesion. We see multiple cystic areas on these T two weighted images in the segmental pattern. There's hypo enhancement here. This represented peribiliary cysts in an area of confluent fibrosis. This is a benign entity, but we might wanna follow those. If we were at all worried LR-2 cirrhosis associated siderotic nodules are nodules that appear slightly distinctive from the remainder of the cirrhotic liver.

For example, these are out of phase T one weighted images and these are in Phase T one weighted images. We see that these lesions here demonstrate some blooming on the higher TE sequences. These are hemosiderin related nodules, siderotic nodules.

Now here's a lesion that if you do MR for cirrhosis, you will come across every single day. It's a NASH-like arterial phase enhancement. This is a lesion that you can see on the early phase enhancement images but has no correlate anywhere else. Has no T two correlate, has no washout, has no diffusion correlate. Most of the time this is gonna be a perfusion abnormality. However, rarely these do turn out to be hepatoma if left to grow. So when we see this, we usually do recommend follow up of these lesions. The vast majority of the time they will be benign.

Suspicious LI-RADS Lesions (LR-3 to LR-5)

Now we will cover the bottom portion of the chart, which is really the crux of the chart that discusses the suspicious RADS lesions. And when we look at these suspicious RADS lesions, we see that we have suspicion categories anywhere from LR-3 to LR-5.

And we are first gonna talk about the major feature of arterial phase hyper enhancement, which is along this axis. It is the single most consistently seen feature in HCC and it is the most important feature of HCC and again, probably due to neovascularity from those unpaired hepatic arteries.

When you see hyper enhancement, it must be unequivocal. And hypo enhancement is much rarer occurring in five to 10% of HCCs. And what's interesting to note is that only arterial enhancing lesions can be categorized as an LR-5, a very suspicious lesion.

So these are examples of different hepatoma that have arterial phase hyper enhancement. These are all very enhancing relative to the adjacent parenchyma and that's a very important point when we're talking about enhancement. We're talking about it relative to the adjacent parenchyma. It's not relative to the lesion itself. These are not time signal intensity curves relating to the lesion. But again, a comparative assessment based on the adjacent parenchyma. All of these are hyper enhancing relative to the adjacent enhancing parenchyma.

Now again, some of these lesions will be enhancing, but there will be a small percentage of hepatocellular carcinomas that do not enhance. We see that here. This was the lesion that underwent biopsy previously as well.

And then when we talk about lesions in terms of arterial phase, hypo or hyper enhancement, we also need to talk about diameter because lesions that are bigger are easier to characterize and perhaps need fewer of these major criteria in order to characterize them. An important note is that if you have a lesion that's less than 10 millimeters, you cannot call that an LR-5. And the feeling is that these lesions are really too small to characterize and we're striving for 100% specificity and we won't necessarily be able to provide that in small lesions.

Now that we've talked about the arterial phase, hyper or hypo enhancement, we will talk about the other major criteria in the LI-RADS chart. And these include washout capsule appearance and threshold growth.

Now, when we talk about washout, that has to do with the drainage pattern of the lesion. And again, hepatoma will show a washout of contrast and that's thought to be due to altered venous drainage and fewer portal venous triads. Again, when you see washout, it must be an unequivocal finding.

These are examples of different hepatoma as they'll wash out. Again, relative to the adjacent hepatic parenchyma capsule appearance is another major criteria and what that is is a peripheral rim of smooth hyper enhancement in the portal venous or delayed phase.

Now this may represent a true or pseudocapsule and the capsular appearance increases with later phases of enhancement. Capsule appearance is a very high positive predictive value for HCC and these are examples of different lesions that have this capsule appearance. A smooth, thin rim of enhancement which increases with subsequent scanning.

And lastly, threshold growth is a major criterion and this is defined as a greater than 50% growth in the largest dimension of the lesion within six months.

Ancillary Features

Now there are other features that we look at. These are called ancillary features that can modify the likelihood of HCC. They can up classify or down classify a lesion based on these features and we can upgrade by one or more categories, but we cannot upgrade beyond an LR-4. We cannot make this an absolute HCC LR-5 without having the major criteria present. And again, we can't do this because we wanna maintain high specificity for HCC.

So some of the ancillary features that favor HCC include mild to moderate T two hyperintensity restricted diffusion, corona enhancement mosaic architecture, nodule in nodule architecture, intralesional fat, lesional iron or fat sparing blood products in the diameter increase less than the threshold growth. And I'll show you some examples of those.

This is a patient who has a hepatoma and this is the inphase on the outphase image. We see that there's some signal loss on the outof phase image suggesting intralesional fat. So this lesion was a fat containing lesion with mild hyperintensity on the T two mosaic enhancement. This is compatible with an HCC.

This is a different patient which demonstrates corona enhancement. Corona enhancement is a irregular rim enhancement. It differs from capsule appearance in that this occurs very early on in enhancement and has an irregular margin. Capsule appearance occurs later on in the more delayed phases and is smoother.

This is a different patient with a very rare finding of a nodule in nodule. This patient had a CT scan and had had this low attenuation lesion that had been stable for years on follow-up. A hyper enhancing nodule was growing within this lesion and this was hepatocellular carcinoma.

Ancillary features that favor malignancy include homogeneous marked T two, hyperintensity or hyperintensity. The presence of normal undistorted vessels coursing through a lesion enhancement that parallels blood pool, progressive enhancement, diameter reduction and diameter stability of over two years.

Examples of LI-RADS Lesions

Now that we have all the information, I'm gonna show you some examples of different lesions on LI-RADS. Here's a patient, who have we seen before? This is a patient who has a 24 millimeter lesion in the left lobe of the liver. Mildly hyperintense T two hypo enhancing on the early and delayed phases, but with restricted diffusion. So we have a relatively large lesion, but it is hypo enhancing. It has some of these features, but again, because it's hypo enhancing, we can never call it an LR-5. This is an LR-4 lesion and this was a lesion that underwent biopsy. In order to obtain proof for HCC.

Here's a different lesion, mildly hyperintense mosaic hyper enhancement on the early phase washout with capsule appearance. This fulfills all the criteria. This is an LR-5 lesion And here's a different lesion. This is a 24 millimeter lesion, hyper enhancing, washout, some capsule appearance. This also is an LR-5 lesion.

Now what about this lesion? This is a small eight millimeter lesion. It definitely has arterial hyper enhancement. We could not say definitively that this was washout, however. So this is a small lesion. It's less than 10 millimeters with enhancement. Without washout, it would be either an LR-3 or LR-4. This patient returned back. This lesion had grown to 11 millimeters. It's now threshold growth, but it had grown and now there is a little bit of washout here. So now this becomes an LR-4 lesion.

Role of Ultrasound in LI-RADS

Okay, so where does ultrasound fit into this classification system? And thanks to the newest version, ultrasound now has a role. And the role is that if you have a lesion that's 10 to 19 millimeters and is an LR-4 lesion, if you are able to see that lesion on antecedent ultrasound, that lesion then becomes an LR-5 lesion and can be treated as a hepatoma.

And let me show you examples of that. Here's that same lesion, the 11 millimeter lesion that had grown slightly. This again is an LR-4 lesion, but we saw this lesion on ultrasound corresponds to this lesion because we can see it on ultrasound. This now becomes an LR-5 lesion. It could be treated as a hepatoma.

Here's a different patient. This is a 12 millimeter peripherally enhancing lesion with some washout. Again, it's an LR-4 lesion, but we are able to see it on ultrasound. It becomes LR-5 and becomes compatible with a hepatocellular carcinoma.

Another scenario where ultrasound can be very helpful is when there are multiple nodules and we're not sure if they're dysplastic or perfusion. The presence of an ultrasound lesion suggests that this is a true lesion and this can be very helpful in this scenario.

So ultrasound has a small but a significant role. And the inclusion of ultrasound unifies the AASLD UNOS OPTN systems. There are several LI-RADS working committees looking at the use of ultrasound and looking at the use of ultrasound contrast. So the role of ultrasound in this classification system will likely expand.

Conclusion

So in conclusion, hepatocellular carcinoma is a worldwide health problem with geographic variations and high mortality. If left untreated, ultrasound is the only imaging screening modality, which has been proven to decrease mortality in large randomized controlled trials. And is the recommended modality for screening by the AASLD.

Ultrasound technique is extremely important to avoid miss cases and the ultrasound appearance of hepatoma is very variable. Findings of HCC on CT/MRI and ultrasound contrast reflect the process of hepatocarcinogenesis, the LI-RADS classification system, although complicated is a precise means of defining and categorizing lesions.

And currently ultrasound has a small but important role in the LI-RADS system and this will likely expand with the ultrasound working groups and the advent of ultrasound contrast. Thank you.