Hepatocellular Carcinoma: Current Imaging Guidelines - HD

Introduction

Thanks very much for the kind invitation to this course.

I am not totally sure I centered the idea, because I was not going to show the findings specifically to CEUS but just to how to incorporate according to the title of this course CEUS into your clinical practice.

The images are clearly important, but clearly in your daily practice, you have to take decisions according to what the scientific bodies recommend, that it is acceptable and is able to be established.

Lesions in Cirrhotic vs. Non-Cirrhotic Livers

We know that for hepatocellular carcinoma, we have a special situation which is different from non-cirrhotic livers because in non-cirrhotic livers, there are a variety of lesions that can be found, and most of them are either metastatic disease or primary benign tumors.

When we switch to cirrhotic livers, which is the setting of patients in which we may detect hepatocellular carcinoma, we have to consider what type of lesions we are going to possibly find, which is different from the non-cirrhotic liver.

Patient Series and Data on Lesion Types

These are some series of patients reported from different countries. And as you can see, all biopsy proven lesions. In this situation, we had about two thirds of HCC and one third of non-HCC lesions, and among them, the large majority are hepatocellular lesions, either with or without dysplasia. But occasionally we also have cholangiocarcinoma or rare tumors like epithelioid hemangioendothelioma.

This is a French study. Clearly, sometimes these nodular cell lesions are only cirrhotic regenerative nodules. So you may find only chronic liver disease.

This is an Italian series from southern Italy covered by fine needle biopsy. And also in this case you see you have other non-hepatocellular lesions like primary non-Hodgkin lymphoma or other hepatocellular lesions.

These are series from three different big centers, all prospectively collected and confirm histologically. And just to summarize, this is our data from prospective studies run in our center. This is from the Barcelona center, and this is from the Milan Center.

Probability of HCC Based on Lesion Size

But just to summarize from the likelihood approach of this study showed that if a lesion is between one and two centimeters, then in one two third of the cases is hepatocellular carcinoma. And as the size increase, the probability the disease hepatocellular carcinoma increases to about 85% between two and three centimeters and 90, 95% if it is more than three centimeters.

But clearly this means that it remains a number of cases which are not HCC. And so if they are not HCC, what do these nodules are? They might be regenerative nodules or without dysplasia, which comprise the large majority of the non-HCC cases in a very few number of cases, which is between two and 3%. There might be intrahepatic cholangiocarcinoma, which is increasing, more often discovered, lymphoma or very rarely metastasis or other extremely rare tumors.

And for unknown reasons, metastasis in cirrhotic livers do harm quite poorly. So it is a rarer occurrence of metastasis than in non-cirrhotic liver.

Evolution of Guidelines for HCC Diagnosis

So having this in mind, we try to look how guidelines incorporate imaging for the diagnosis of HCC. And this is quite new because it is the only tumor in which we can establish a cellular definition of the tumor of the cancer, not only that it is malignant. For instance, for metastasis, we can say this is a malignant lesion, but if we do not know the primary tumor, we are not able to distinguish breast or colon metastasis. So we can only state this is a malignancy. But for hepatocellular carcinoma, we are allowed to establish a diagnosis of the type of cellular of cells, which is hepatocellular lesion.

So this has been accepted only in the last 15 years, since the publishing of this consensus conference, which adopted for the first time noninvasive criteria and accepted that if a liver nodule more than two centimeters shows the presence of arterial hypervascularity. And at that time, the CEUS did not exist at all. So they just adopted CT MRI and angiography, which is now abandoned or doppler ultrasound or one, either two of these imaging tools or one of these imaging tool plus alpha fetoprotein. Then the lesions could be accepted as an HCC.

So this was really a striking innovation because up to that time biopsy was always needed. And clearly biopsy is not always feasible in cirrhotic patients.

2005 Update in American Guidelines

An update came out in 2005, so a few years later after the release of CEUS, and it did quite interestingly for us, although these were the American guidelines, the CEUS was endorsed, as you can see here, together with CT scan or MRI and update of the criteria was introduced that not only hyper vascularization was enough to establish diagnosis, but also the washout was required.

And that was not because of CEUS because with ultrasound we always target the lesion. But because with CT, often you can see hyper enhancing lesion, which are only small fistulas. And clearly if you accept only the previous criteria, this could be defined as tumor, whereas they are not tumors, but they are only intrahepatic fistulas.

So in order to avoid this risk of misdiagnosis, they required that the suspected region should also have a wash out, so to appear hyper enhancing in the portal and venous phase. And together with the introduction of this pattern for more specificity, the role of oncofetal markers was abandoned.

So that was quite clear, and we could just stay finish here because all techniques are suitable and able to establish a diagnosis of HCC.

2010 Article on Misdiagnosis Risk with Cholangiocarcinoma

But a few years later, in 2010, this article appeared from the group of Barcelona showing that a limited number of cases belonging to cholangiocarcinoma showed the same pattern of hepatocellular carcinoma. On contrast enhanced ultrasound, which means that these 21 lesions, which I would like to outline, were collected over six years. So it means one to two patients per year in a big center, which is a very limited number of cases. But nonetheless, this is an important topic to touch.

Out of this 21, a bit less than half showed the homogeneous hyper enhancement in the arterial phase, which is typical for HCC, followed by washout hyper enhancing in the venous and delayed phase. And this clearly has the same pattern that has been accepted for HCC. So this is potentially a risk of misdiagnosis.

I would like also to outline that MRI is not able to establish a diagnosis of cholangio. So this means that we pay the risk of misdiagnosis with the reality of not being able to establish a diagnosis. And clearly these patients are to be referred to biopsy, which is correct, but biopsy unfortunately cannot be always performed.

What I would like to outline that was not mentioned specifically in this paper at that time, was that a large number of cases already washed out in the portal phase here, it was defined up to two minutes, so we do not know whether it was 40 seconds or 90 seconds, but the majority had a relatively early wash out. And this may have some relevance.

The remaining cases had the rim enhancement, which is not typical for HCC, and so it is not accepted for the noninvasive diagnostic of HCC.

Confirmation Studies on Cholangiocarcinoma Patterns

This data were confirmed by study that we set up together with our colleagues in Milan, and that we publish in liver international are confirming that a number of cases which are here are at risk of this diagnosis, because this cholangio were hyper enhancing in the arterial phase homogeneously and were washing out. And as you can see also in our series, more than half of the cases were already hypoenhancing in the portal phase.

And as confirmed, not all, but most cases at MRI did not show washout. So they are hyper or the enhancement did only a part of the tumor and not the entire tumor.

So because of these potential risk of misdiagnosis, the guidelines were updated. And in the latest release of the American guidelines, as you can see here, there is no more CEUS. So what the American guidelines allow us for establishing a diagnosis of HCC, they only allow CT or MRI. And what is of relevance is that this flowchart allows only the diagnosis of HCC. So if a lesion is not definitively an HCC, it should be biopsied.

Clearly this is very scientifically sound, but in real practice, not all lesions can be biopsied. So in the lesions that cannot be reached because of clotting disorders, ascites, unsafe needle track, we do not have the possibility to reach a diagnosis or not even the risk that this lesion is an HCC.

Differences in International Guidelines

So as we already outlined, there are differences in the guidelines. I have showed you the American ones because they were the first to endorse CEUS, but clearly according to the practice in different parts of the world, and according to the use of different contrast agent with the different techniques, there are differences in the criteria adopted to establish a diagnosis of HCC.

And just to summarize, the American and European guidelines released by the Hepatology Society do not endorse CEUS anymore since 2011, 2010. Whereas the other guidelines, including the Asia Pacific, the Japanese, the World and European Federation of Societies for ultrasound in medicine, biology, and the Italian Association for the study of the liver, they still do endorse despite these findings, the use of CEUS as they will illustrate.

European Guidelines

Now. So these are the European guidelines of the European Society for the study of the liver, and clearly they adopt the same criteria. So the diagnosis of HCC is based on noninvasive criteria showing wash in and wash out, and they adopt only CT or MRI. So they have the same strategy as the American guidelines differently in Asia Pacific.

Asia Pacific Guidelines

The strategy to come to a diagnosis of liver nodules in cirrhosis is different. And I would like just to remark that in patient at risk for HCC, we include all cirrhotic patients, regardless of the etiology. And for hepatitis B, also patients male, older than 50, 40 years of age with F two or F three fibrosis with moderate or severe fibrosis, because these patients are at increased risk of hepatocellular carcinoma.

So these are the categories in which we may establish diagnosis of HCC. When we want to characterize a lesion on this background setting, I would like to comment the different techniques. And as you can see here, when they released this guidelines, some of CEUS was not yet register in these Asian countries, and they adopted two techniques that are different from the other guidelines.

One is the super paramagnetic iron oxide contrast agents for MRI or the Kupffer phase on gadoxetate or levovist. So the late uptake of the contrast by Kupffer cells, which clearly lack in malignant lesions. So this guidelines start from the pattern in the arterial phase. If the lesion is hypervascular, RCT at CT or MRI, then you see what happens in the portal phase. And if it wash out similarly to the European and the American guidelines, you can establish a diagnosis of HCC quite differently from the European and American guidelines.

If there is no washout, then you can adopt additional techniques, which are the ones mentioned here. And if there is a lack of uptake, which means lack of Kupffer cells combined with the hyper vascularity in the arterial phase, they accept the diagnosis of HCC, otherwise biopsy are needed or close follow up.

These guidelines accept CEUS as I have already shown among the techniques that can establish the of HCC. And the criteria are the same arterial enhancement with portal vein washout and alpha fetoprotein was abandoned as well, because this is not specific enough. It may raise also in case of flare up of hepatitis and other situation, regenerative situations. So it is not acceptable.

But they have also a flow chart for lesion which are hypovascular in their arterial phase. Hypovascular means lacking hyper vascularity. So both the lesion which are iso enhancing or hypoenhancing in the arterial phase are included here. And they accept, according to what they say here, when there is a Kupffer phase defect or defect of the super paramagnetic iron oxide uptake, they accept a diagnosis of HCC even in the absence of hyper vascularity.

I am showing it to you, but also criticize it because this is not validated. And clearly the risk of misdiagnosis with other malignancy is very high because also other malignancy do not have Kupffer cells, the lack arterial hypervascularity. So I do not like very much this guidelines. Also, I like to show them because they have at least one important new concept that they say CEUS is more sensitive for detecting arterial vascularity of target nodules than dynamic CT or dynamic MRI.

Okay, we know and spec radiation ve nicely how the real time allows to pick up arterial hyper vascularity, which is even very transient, only for a few seconds, but they conclude. Thus, hypervascular hypoechoic nodules on dynamic CT may be diagnosed by CEUS.

So this is sort of new modality, which is called the integrated imaging. We may put together some features or one modality together with some feature of another modality. This is a very tricky situation because clearly it is, it depends on what you combine together, because combining a face of a person with the legs of another person, you might create a wonderful girl, but also a terrible monster if the combination is not very good.

So it is important that what is useful is only to pick up the arterial hyper vascularity, because clearly if we pick up the wash out of CEUS early and late with the hyper vascularity of MRI, maybe we are diagnosing only cholangiocarcinoma. But this is interesting because they accept the integrated imaging. So this is something which the other guidelines have not endorsed yet, but I find it very interesting.

Japanese Guidelines

What is the situation in the Japanese guidelines? These guidelines are produced by a nation where the number of HCC is very high. So they have a lot of knowledge. Sometimes they are a bit complex when they make things, so it is not very easy to follow. And again, they start from the vascularity in the arterial phase. And as you can see, these guidelines are from 2011. They had already experienced quite an experience with contrast enhanced ultrasound, and they accept any technique. And if there is hyper enhancement and washout clearly, we have the typical HCC.

So all these guidelines still endorse the use of CEUS with washout for diagnosis of HCC. So they are different from the European and the American guidelines because they dropped it only because the risk of misdiagnosis with cholangio. And I will touch this point later, but I would like to remind that this occurs only in very limited number of cases.

But very interestingly, they endorse some new features, which is the use of hepatobiliary phase contrast agents at MRI. So I am trying to make an overview of the different guidelines, because as you can see, there are differences that are interesting for knowing how things are moving on. There is not only CT or CEUS, but there are different contrast agents and the possibility of integrated imaging.

And they adopt the hepatobiliary phase contrast agents, which provide any information about the function of cells of hepatocellular cells. And if there is a defect in the post vascular phase, which means that these cells have not taken up this type of contrast, which means that usually there are tumoral, then if this combined with hyper vascularity in the arterial phase, you can establish a diagnosis of HCC.

Clearly they also endorse the possibility of integrated imaging. So if you do not see the arterial phase hyper enhancing with CT or MRI, you can add CEUS. And if you show hyper enhancement, you define this lesion as hyper enhancing.

This is quite important because clearly sometimes you do not see the hyper enhancement because it is not there with CT or MRI, you just do not pick it. But quite often, as you probably are aware, there are technical issues. For instance, the patient is breathless during the contrast infusion of MRI. And you cannot repeat a new MRI on the next day just because the arterial phase was not satisfactory. And this, it is also told that happens more often with the hepatobiliary contrast phase agents. This is not known, but the patients become less able to keep the breath inside probably is a chemical reaction. It is not known. So the rate of poor function of MRI in the arterial phase is a bit higher with these contrast agents.

So they provide more information about the post vascular phase, but they might be technically insufficient in the arterial phase. So it might become even more important to combine together the information provided by MRI but with poor technical arterial phase with CEUS, which is very good to provide information about the arterial phase.

So this is the flow chart for hypo hypervascular lesion. So lesion which do not have the hyper vascularity in the arterial phase, and they also use the possibility of these post vascular agents or Kupffer phase gadoxetate, which is contrast agent released only in Japan, China, but not in the European community. And according to these, they combine these different parameters. They have this threshold of 1.5 centimeter, and this has been quite complex.

So they updated these algorithm recently. They made it more simple and they start from dynamic CT or MRI, which is also what is reasonable because clearly you need to stage patients, so not only to define these lesions, but to see if there are extrahepatic metastasis and the lymph node if there are other lesions. So CT and MRI are needed from a cost effective point of view. It is reasonable to start from CT or MRI, but this does not have to exclude CEUS as the American and the EASL guidelines did.

CEUS can clearly utilized in many situations, like for instance, also patient with contraindications, because if you do not allow the diagnosis, then you cannot make the diagnosis even in patients who cannot undergo CT or MRI. So I think these guidelines are much more reasonable. They accept CEUS, they made it simpler hyper enhancement, plus washout HCC. If there is no washout and the lesion is very small, then just follow up. If the lesion is more than one centimeter, you include optional testing and optional testing means CEUS and biopsy when needed. So it is accepted and it is endorsed here in this box.

UK and Italian Guidelines

What are the guidelines in UK? In UK also for the use of CEUS, there is acceptance of the method to establish a diagnosis. And clearly it is utilized when enhanced ultrasound is inconclusive. But this is nearly, let us say 100% of the cases in HCC, unless the tumor clearly invades a vein. So you have already diagnosis of malignancy, but for small nodules, it is always needed. And when MRI is not clinically appropriate, not accessible or not acceptable to the person. So this is accepted.

And in Italy as well, we accept CEUS, because clearly I have already mentioned how it is able to make a diagnosis, but also because in Italy we got quite a large experience in CEUS so we already refined our criteria and one of these important criteria is that the global enhancement, which is typical for HCC, so the entire nodule enhances in the arterial phase and wash out in the late phase as it has already been told this morning.

The wash out of HCC typically is mild. It does not become a black hole. It is less enhancing than the surrounding parenchyma. And usually this washout starts after two minutes, and one fourth of the cases do only wash out after three minutes. So one important recommendation for CEUS in cirrhotic livers to characterize the lesion is to scan the lesions up to four or five minutes. Clearly this depends on the amount of contrast that you have injected on your equipment and how sensitive.

And very importantly, it is important not to keep on scanning, scanning, scanning because you disrupt the microbubbles. It has been shown already this morning. And clearly these reduce the number of bubbles that are present after three minutes, and you might not be able to use this information. So after the first 40, 45 seconds or one minute, you can stop scanning and you start scanning intermittently every 15 to 30 seconds, you stay on the lesion and you switch on just for a few seconds just to understand whether the lesions remain iso enhancing for the entire life of the microbubbles or wash out lately.

However, in cases they wash out early and early means less than 60 seconds. This is not typically for HCC, and this does not mean that it is not an HCC, but in case it is an HCC, it is very poorly differentiated. Or quite often it is a non hepatocellular malignancy. So it might be a cholangiocarcinoma in first instance or more readily metastasis or lymphoma.

So it is very important to assess these lesions continuously for the first minute because this is a key situation.

So the Italian position is that clearly CT or MRI are important for staging, and if a patient is seen in the community center and it is detected with a lesion, probably it is better to go at first to CT or MRI. But CEUS is clearly entirely accepted first because it is able to define whether lesion is malignant or not, because clearly it might be hepatocellular carcinoma or cholangiocarcinoma, but both entities are well recognized as malignant, whereas MRI is able to establish a diagnosis of HCC as malignant lesions. But in case of cholangiocarcinoma is not able to distinguish benign from malignant lesions because cholangiocarcinoma do not wash out.

So the information that gives tells us that this is a malignant. So this is why it is for US Italian, for many other societies that are shown to keep CEUS, because at least to a patient that is not able to receive a biopsy, you are able to distinguish between benign and malignant, which is extremely important.

And there are some situation where it is contraindicated or might utilized as a second instance. Clearly, if a patient is detected with a new lesion in a center, which is able to immediately perform CEUS, this is what we do in daily practice because clearly if you prescribe an MRI, the patient might wait one month, but if with CEUS you can immediately know whether the lesion is malignant or not, and whether it is an HCC or not, depending on the timing on the washout, you can clearly, you still need a CT or MRI to make the staging, but you can immediately plan the treatments, start to discuss the situation with the surgeon, with the radiologist. And so this is what is recommended. It at least in Italy, the FISM and FADOI guidelines also devoted a chapter to this.

And in this section we already distinguished different situation at risk. And I think this was sort of a good precursor of the VIRADS that professor Cosgrove already illustrated. And it is sort of innovation because up to now the guidelines on the stated this in non-HCC, okay, yes, no, it is not, you need a biopsy. But these guidelines do not help us to tell how to manage patient who cannot receive a biopsy. Because as I mentioned it, and you are aware, many of these patients cannot receive a biopsy. The lesion cannot be reached. So there are multiple lesion. We cannot make three biopsy or four biopsies in different nodules.

So apart from confirmation of the pattern, so arterial hyper vascularity and if it wash out in the late phase and is mild, is malignant and considering HCC, unless it is early and very strong washout, but if there is no enhancement, we are not able to make definite diagnosis, but at least there are some intermediate risk of being cancer.

If it is no hyper enhancement and it is iso enhancing, no washout, then this is indeterminate for malignancy. But if there is no hyper enhancement, but there is washout, then there risk of HCC is bigger. So it is indeterminate, but highly suspicious for HCC. And if there is no hyper enhancement but very marked and rapid washout, then it is very suspicious for malignancy. And we have also to consider non-HCC.

Limits of Current Guidelines and LI-RADS

So to summarize and go into the last few slides, CEUS was accepted in the American guidelines of 2005, but not included in the following guidelines, even from EASL because of the misdiagnosis with cholangio, which however I showed you it is very limited and with CEUS it can be distinguished from HCC in about 80% of the cases according to the timing and intensity of washout.

And we had a paper in ultrasound last year from China, which clearly demonstrated the timing in the median timing of washout of cholangio in cirrhosis is less than 60 seconds. So this is very important. And the pattern hyper hypo, despite the risk of general malignancy, if we do not consider timing is a bit specific for malignancy.

What are the limits of the current guidelines? They recommend only how to establish a diagnosis of HCC, but they do not tell us how to manage lesions, which are not HCC and cannot be biopsied. And another problem with which is we still see is that the modality for reporting the radiologic findings is quite variable in my experience sometimes is exotic, artistic. You see the very best description of this nodule T2, T1, and then you do not have any conclusion. And for as a clinician, this is worthless. We have sent the patient for an examination and we do not get any information.

So to overcome these problems, the American College of Radiology set up working groups, different working groups which deliver recommendation how to report and design the spectrum of probability of lesions to be HCC and to be benign versus malignant. This maybe seems very basic, sometimes quite stupid, but it is very important that all radiologists or imagers provides the conclusions in the same mode.

So this is the work that has been provided for CT and MRI. They started with CT MRI and you can see that there is a spectrum of lesions going from LR one, which are definitively benign lesions, hemangioma simple cyst, probable benign, which is something that for CEUS to me is more difficult to define because here they include for instance, THAD, which are fistulas that we do not have in CEUS. So clearly this section is much more limited for CEUS.

And if it is not definitively or probably benign and it is not other malignancy, then we have different probability of hepatocellular carcinoma. Hepatocellular carcinoma is LI-RADS 5. LI-RADS 4 is highly probable hepatocellular carcinoma, although not definitively hepatocellular carcinoma and LI-RADS 3 is intermediate risk of hepatocellular carcinoma, whereas LI-RADS 5 is tumor in vein vascular infiltration as it was anticipated the CEUS LI-RADS scheme has been under production.

I am not going to present it because it is not been finally accepted and is not released. So this is just a draft probably would be the final version, but it is not online yet. Hopefully it will be within the summertime, but clearly I am not allowed to present all the information. But as you can see, we kept the same strategy and we have as definitively benign, only clearly cyst hemangioma focal fat infiltration of focal fatty sparing as probably benign. There is very, very limited number of cases, very small nodules, less than 10 millimeters, iso or iso enhancing.

And what is of interest is that we can make a sort of probability risk of being hepatocellular carcinoma. This does not, in my view, does not spare biopsy. Biopsy is still needed, but at least helps in all the cases in which biopsy cannot be performed, which is not so rare. So I hope that this strategy, which up to now has been not validated and the draft said will probably become more spread everywhere.

Summary and Conclusion

So to summarize the diagnostic criteria, the type of contrast agents and imaging techniques differ in different geographical areas and related guidelines for the diagnosis of HCC. Nonetheless, all guidelines accept a noninvasive diagnosis of HCC, at least by CT or MRI when showing wash in and washout. Very few adopt additional nonvascular criteria like the Kupffer phase or hepatobiliary phase uptake by MRI contrast agents, but their specificity is still questioned.

I did not go into detail, but for instance, the hepatobiliary phase uptake defect is seen also in FNH nodules. So by itself is not enough to define a nodule malignant integrated imaging is still not widely endorsed, but I think this is one of the most important key features for the future.

And most guidelines accept also CEUS for lesions in cirrhosis. And to establish a diagnosis of HCC or for CEUS is adopted in second line. Because CT and MRI are requested for staging, but not because they are better at characterizing the specific lesion for the specific characterization CEUS in most instances of very well visible lesions is probably superior.

And caution is recommended when washout occurs early. I mean caution in making a diagnosis of HCC when washout occurs early less than 60 seconds and is very marked and new standardized criteria encompassing the entire spectrum of lesions in patients at risk of HCC have been proposed but are not yet fully validated nor widely endorsed, but probably this is only a matter of time.

Thank you very much for your attention.