Andrea Branch, PhD , Professor of Medicine, Division of Liver Diseases, explains the global importance of hepatocellular carcinoma (HCC), the predominant histological type of liver cancer; describes liver cancer screening guidelines and demonstrates how they may disadvantage Black patients; and identifies barriers to optimal liver cancer screening that impacts all patients.
Referring a patient is easy. Just click the “Refer a Patient Online” button. DR branch. Good morning everyone. Welcome to grand rounds. Um it's my pleasure to introduce dr Andrew branch today she received her undergraduate degree in zoology at the University of michigan. Her masters in cellular and developmental biology at UC Irvine and completed her PhD and postdoc in molecular genetics and addictive diseases at the Rockefeller University. She's a professor of medicine and director of basic and translational research, a Mount Sinai's institute of liver medicine and her research focuses on viral hepatitis including hepatitis C and HIV the patio cellular carcinoma and the immuno biology of the human liver. Using both basic and translational approaches, her lab has published numerous papers examining clinical outcomes of hepatitis C treatments, viral and human genomics and liver diseases and risk factors and detection of the patio cellular carcinomas. She has a long standing interest in environmental health and has studied the risks of environmental exposures such as proximity to superfund toxic police sites and the World Trade Center as they relate to the development of liver disease. A very warm welcome to DR branch this morning. Uh thank you very much for that kind introduction. So my truck today really features some work which is done been done over the years by three of the residents here at Mount Sinai. So I thought that might be interesting. And it really speaks to a problem that we found in which it seems that african americans who have a high risk for developing liver cancer don't really fit the profile, which was largely developed by examining caucasians who were at risk for liver cancer. And we're concerned that this disconnection between the presentation of liver cancer and african americans um causes them to not really fit the clinical image of what a high risk person looks like. And this could be disadvantaging them. So these are my disclosures. So the objectives are to describe the importance of liver cancer, which is the predominant histological type of liver cancer, to describe screening guidelines and then to highlight some of the barriers to liver cancer screening. So for the background we'll talk about HCC um cirrhosis being a prerequisite for HCC surveillance. This is recent data showing the global cancers and you can see that liver cancer is the fourth leading cause of cancer related deaths globally. And really the number of deaths are very similar to stomach cancer and colorectal cancer and then considerably higher than breast cancer. So despite this very high global mortality rate, unfortunately, um survival following a diagnosis of liver cancer is very low. So survival is down with pancreatic cancer, lung cancer and cervical cancer. And survival. After diagnosis of liver cancer unfortunately has really been stuck at a very low level and has not progressed as much as we would like over the last five years or so. The guidelines for liver cancer surveillance are heavily dependent on the presence of a sort of precursor liver lesion, which is liver cirrhosis. So when you take a normal lover and you expose it to liver damage over the course of many years, you develop scar and cirrhosis, And once cirrhosis develops, a patient has an annual risk of developing liver cancer. That's about 1%. About 80-90% of all Japan cellular carcinoma arises in a psoriatic liver, but about 10-20% um, does not pass through this state of cirrhosis. The rate of progression from cirrhosis of the liver cancer varies considerably depending upon the underlying liver disease that causes cirrhosis in the first place. So for patients with non alcoholic fatty liver disease and alcohol is their exposures, it's less than 1%. In the case of patients with chronic hepatitis C is the underlying liver disease, it's almost 3%. But regardless of these numbers, whether it's less than 1% or up to 3%,, these are very high incidence rates compared to incidents for other cancers that trigger screening. For example, Rescue The incidence of breast cancer in women over the age of 50 who were advised to have annual mammograms Is only 0.35%. And here's the incidence of colon cancer in individuals who are advised to have colorectal screening. So you can see that the barrier to qualifying for HCC surveillance is very high compared to other answers. However, it seems that uh can cordant adherent surveillance is very important for early diagnosis and effective treatment of liver cancer. This is a study that was carried out in France and it compared lead time adjusted survival and so erotic patients depending upon whether they had liver imaging at intervals of less than seven months or greater than seven months. So it appears that it's very important not only to participate into surveillance but to Have surveillance at approximately six month intervals. Despite this, the National Cancer Institute does not recommend liver cancer surveillance at all. And the United States preventive Services Task force also does not recommend liver cancer surveillance and that is because they do not feel that the benefits outweigh the risk. So I think it's really important to consider what keeps the entry barrier to liver cancer surveillance so high. And I see this as a series of Catch 22. So the preventive task force and N. C. I do not recommend surveillance because they say that the evidence in favour is not strong enough and we would probably require a randomized controlled trial to change their position, but but clinicians and patients would not agree to be enrolled in the trial. So we've got this sort of standoff. So the American Association for the Study of liver disease recommends twice annual ultrasound for patients with cirrhosis. But here the problem is that it's really hard on patients and providers to have twice annual surveillance. And besides that ultrasound is not really a very sensitive modality. There are more sensitive procedures of CT and MRI but they have their own problems cost and in the case of C. T x ray exposure, in the case of MRI, somewhat limited availability. Um And a final catches that practice guidelines really have only limited power of enforcement. So we can certainly uh encourage providers to adhere to our practice guidelines. But they just do not have the force uh that a statement from the US preventive task force would have. So if we had better survival of our liver cancer patients that would increase the life years gained and lower the incidents at which surveillance would be considered cost effective. But here the catch is that to get the improved survival, we need an increase in early detection in order to get early detection. We really need more surveillance. So we're in a bit of a bind. Um This slide just uh enunciates what the A. S. L. V. Guidelines are for adult patients with cirrhosis surveillance as advised twice annual ultrasound with her without AFP for patients who have been cured of hepatitis C. If they had cirrhosis at the time they were cured still advised to undergo twice annual imaging. However, they feel that the risk is enough lower in patients without cirrhosis that it is specifically not recommended for these patients. So surveillance is recommended for selected groups of non cimirotic hepatitis B infected patients. But this expanded guidelines does not extend to US born african american. The V. A. Guidelines are a little bit broader and more nuanced than a SlB guidelines. So they expand down one stage stage three as well as stage four fibrosis, which defines cirrhosis and acknowledge that some patients with chronic hepatitis beer at elevated risk and may benefit from surveillance. This slide is a reminder that once a patient arrives at cirrhosis which we would also call stage for f for fibrosis. This is a histological definition but it actually spans quite a range of clinical presentations all the way from a well compensated patient with no viruses to uh some compensated cirrhosis and then finally decompensation with the sides and hepatic encephalopathy. These are some of the clinical features of advanced liver disease. So you have a nodule our liver, you have portal hypertension that results in societies. And you have a stoppage all viruses. So now I want to turn to the distinctive features of liver cancer and african americans that they present with less advanced liver disease and more aggressive cancer. We're testing the hypothesis that african americans do not fit the conventional HCC risk profile because they have an attenuated fiber optic response. With fewer indications of advanced disease with higher platelets, less societies, less liver, natural clarity on imaging when they develop HCC than whites. Uh the significance is that this reduced liver disease progression might contribute to closer low surveillance rates and late detection. So the initial study that we did was spearheaded by Adam Winters who was a resident at the time, and Adam has now rejoined Mount Sinai as an advanced technology fellow and he recruited one of his co residents, julie son, who then joined the project and carried it out with Funny Perumal swamy, who is my colleague here in liver until she Lear uh left for a position of michigan earlier this year. So Adam and julie combined on a project indicating that african americans do have left liver disease when they develop hepatitis C. Um HCC focusing on patients with hepatitis C is the underlying liver disease. And interestingly both Adam and julie subsequently won the emerging liver Scholars award for this project in subsequent years. So why were we focusing on hepatitis C? Well, hepatitis C is the predominant underlying liver disease in liver cancer patients in the United States. So this is work from texas and it's showing different time periods, But you can see going back to around 2008 and up until 2016, about 66% of HCC in the United States had hepatitis C is the underlying cause. Um and this is a study that looks at the relationship between the underlying cause and progression to uh, HCC. And what it shows is that among patients who have cirrhosis patients whose underlying liver disease is hepatitis C have about a fourfold greater incidents of developing liver cancer than patients who have cirrhosis due to other ideologies, such as alcohol exposure or fatty liver disease. And not only is the incidents higher. Um, also mortality is higher. So this is a slide that looks at HCC age adjusted mortality and what it shows is that by far the leading underlying liver disease for HCC mortality in the United States, this hepatitis C infection, And why are we focusing on african americans? Well, that's because african americans have about twice the age adjusted incidence of liver cancer compared to whites. And furthermore, in one of the very, very few prospective studies, looking at the incidents of liver cancer and looking at risk factors for liver cancer Found that African Americans had about a 2-fold higher hazard ratio than caucasians. And I think I put special weight on this study because it was a prospective study in one unusual characteristic about it is that all of the individuals in the study had the same surveillance. So, since detection, early detection certainly depends on uh imaging. If you try to determine HCC incidents rates without ensuring that all participants, all high risk patients have the opportunity to undergo surveillance. You're very likely to underestimate the incidents in populations that do not have a high level of surveillance. And finally, african americans who developed liver cancer have the lowest overall survival. And this just is occurring on the backdrop of the reduced life expectancy. Um that has been seen in african americans for many years compared to whites and greatly exacerbated because of covid. So um along with Italy shuttle and UMA czar people, I did a follow up study along with our colleagues where we looked at an expanded population of african americans with hepatitis C is the underlying liver disease who are diagnosed with HCC. So we had the entire group of patients and then we did two subgroup analysis Our methods. It was a single institution retrospective study of records from 2003 to 2018 included all patients with manual review to confirm HCC diagnosis. FC exposure was defined by anti by your HEp C. RNA chronic, ongoing. The infection was the presence of healthy surface antigen or heavy DNA. Um but importantly, we also looked at Hepatitis B exposure um in the United States where a lot of hepatitis B is transmitted horizontally, as opposed to in asIA where most B transmission is vertical, there's a high rate of clearance of hepatitis B, but it's really clearance is not complete in many cases and you have residual have the DNA lingering in the nucleus and you have a persistence of hepatitis B core antibody positivity. And so even in patients who have lost hepatitis B antigen or hepatitis B. D. N. A, you can still identify patients who had hepatitis B infection at some point in their life by monitoring for hepatitis B core antibodies. We kept track of HIV infection, race and ethnicity were self identified imaging data were collected from reports and the modalities for diagnosis included CT and MRI. We use the Barcelona Clinic liver cancer stage and looked at milan criteria and use the Medicare system for classifying fibrosis and inflammation and tumors were staged according to the american Joint Committee on Cancer. So the study had over 1000 patients, about a third were black and a third or other. And these are their characteristics. So important things to notice are the reduced percentage of african americans who had commercial insurance, um and the high rate Of hepatitis b exposure actually in both populations, but 60%, almost 60% of the African Americans with hepatitis C infection as the main underlying liver disease had a history of hepatitis B exposure. And this is likely to be important from the standpoint of establishing liver cancer risk because as outlined in this diagram of the Hepatitis B lifecycle, although hepatitis B um does not require insertion into the chromosome of Hepatitis B DNA. So, in that sense, it differs from HIV, which inserts in the host chromosomes, is an obligatory step of its replication cycle. Hepatitis B DNA replicates in the nucleus as an episode. But since infection continues for such long periods, pieces of hepatitis B DNA often end up getting inserted into chromosomes. And also you can have very stable populations of the episode. No hepatitis B. DNA and whether it's the episode or the insertion into the chromosome, uh there is an association between exposure to hepatitis B and HCC that we're trying to drill down on a little bit. Um The black patients had much less advanced liver disease by a whole variety of metrics lower billy Rubin, lower I Nr. So better clouding considerably lower, considerably higher platelets. And I make this particular point about platelets because um the platelet count is a very well recognized indicator of liver cancer risk in patients with cirrhosis. So being a population um that has high platelets um is is going to make you seem to be a patient whose liver cancer risk may not be so concerning um to you or your provider. Um on imaging, black patients had much less advanced liver disease, so they were less likely to have Nigel. Our livers. Um the pathology report was less likely to mention cirrhosis, um dramatically less likely to have a cities, so 80% did not have a cities less fundamentally, only half less than half the incidents of viruses and less portal hypertension. So by all these measures less advanced liver disease and yet they had much worse liver cancer. So their tumors were larger, they were more likely to be multiple, they were more likely to have growth, vascular invasion and metastases. And they were less likely to be within the Milan criteria, which are listed here. And the reason this is so important, whether you're within or beyond Milan criteria is because being within Milan criteria gives you a much better disease free and recurrence free survival. If you go under, if you undergo liver transplantation, um within milan criteria, there's a high likelihood that the cancer is actually confined to the liver. So removing the organ is curative and outside Milan criteria, you're more likely to have a tumor outside the liver that than gets back into the graft and reduces survival if you do undergo liver transplantation. So we drill down a bit on the patients who had surgery for treatment of their liver cancer surgery, being either resection or liver transplantation, we did not find any difference in the likelihood that blacks are non black patients would undergo surgery. Um One thing that is advantageous from the research perspective is that when patients have surgery, it gives actual liver tissue that's available for analysis. Um, uh, whereas often we don't have tissue from our liver cancer patient. Um, what we were able to see is that in the liver that surrounded the tumor, african americans had much less fibrosis. So compared to the Non black patients about who had stage four fibrosis or cirrhosis, about 80% of the time. Um, only 64% of the African Americans. And I would just mention one of the reasons why these numbers in both cases are as low as they are, is because if you have advanced cirrhosis, you're less likely to be a resection candidate because you don't have sufficient delivery service for resection. So, the surgical patients are going to be selected for less advanced liberties, generally speaking. Um, so in the patients who did undergo surgery, the african americans had again worst cancer with more micro vascular invasion. Satellite lesions, 30% had poorly differentiated tumors, much less likely to have stage one. The answer And their overall five year survival was less. So given our concern that the higher prevalence of hepatitis B exposure might be skewing the results, we did a subgroup analysis of patients who had no hepatitis B and no HIV, and we found essentially the same thing in this group of hepatitis C mono infected patients, which is to say, once again, they had less likely to have commercial insurance, um, considerably less advanced liver disease, but they were again less likely to be within milan criteria and more likely to have metastatic disease Among the patients who had surgery. Again, the cancer was worse, less likely to have stage one more likely to have micro vascular invasion. And we then did a final combined study where we put both groups together A lot, which gave us over 1500 patients. And not unexpectedly, we found a less commercial insurance as one of the main differences between the black and non black patients. Again, the african american patients had a much more likely to have childs a cirrhosis, the initial uh phase of cirrhosis. Um Bye billy Rubin, white words clotting all indications that african americans had less advanced liver disease and again, they had fewer abnormalities on imaging and yet they had more advanced liver disease with larger tumors, more multifocal tumors, more gross invasion, more likely to have metastatic disease, less likely to be within milan criteria. Um And among those who had surgery again less likely to have cirrhosis, so less advanced liver disease, but worst liver cancer With 30% having poor differentiation. So, um these findings are consistent across studies, so we're not the only group that has seen this. Um It has been found by several groups that african americans have lower billy Rubin and higher platelets. Um So it does not seem to just be a mount Sinai phenomenon. Rather, it seems to be a general feature uh african americans at the time that they present with liver cancer. Their liver disease progression is much less and so one might find less concerning, but they have more aggressive and worse liver cancer. So the conclusions are that liver cancer has a distinctive presentation in blacks with less fibrosis in the surrounding liver, higher platelets, less societies, but more aggressive tumors. Um Certainly we find a lack of commercial insurance and we suspect that this may be both contributing to the challenge that black patients may have in accessing some of the more expensive imaging modalities um and also may just be a general marker for socioeconomic barriers that they face. In any case, HCC screening programs that require cirrhosis made disadvantaged blacks because they do have a much better preserved liver function. And even if they do have cirrhosis, they may not look as concerning as other patients. I'm impressed by the high prevalence of hepatitis B exposure that we found in our patients and feel that this merits reexamination of a HCC risk factor, particularly in african americans. The field has gone back and forth on whether um, core patients who are simply hepatitis B core antibody positive are a high risk group for liver cancer. Um, in studies that have been done in ASIA and in europe, the answer has been yes. In the large healthy study that I mentioned is our one really excellent perspective study that was done in the United States. The court antibody positivity was not found to be a risk factor, but that was a relatively small study and in the United States certainly the prevalence of hepatitis B infection varies quite a bit regionally and we may have a higher prevalence of hepatitis B here in the new york metropolitan area. And so that would allow it to emerge as a greater risk factor for our patients. Um as a piece of kind of interesting underlying biology, it seems that african americans produce less scar in response to liver injury than caucasians and that they may have distinctive liver cancer risk factors that really needs to be teased out and delineated. So I wanted to turn to some information about hepatitis C surveillance and review a bit about the recommended detection methods and some of their shortcomings and some information about the rates of guideline concordance surveillance, which tend to be very low. So there are many barriers to effective HCC surveillance. So I've mentioned that the A. S. L. D. Screening guidelines advise twice annual ultrasound um With or without alpha feta protein measurements for patients with cirrhosis. A problem is that ultrasound is sub optimal insensitivity especially for tumors that are less than two cm. And the sensitivity can be reduced by cirrhosis. Which is obviously a problem. Since it's only recommended for patients who have cirrhosis and also ultra obesity, which is a ever increasing condition in the United States can be a challenge for ultrasound examinations. Alfa feta protein is our most longstanding serum indicator of H. C. C. Um But it has poor sensitivity and the cut points vary across patient groups. So implementation has been poor. So as a result of this um there's been a tremendous rise in skepticism among providers in the United States and about 60% of them do not adhere to the H. The A. S. L. D. Screening guidelines to use ultrasound alone for surveillance. And rather 60% of them combined ultrasound with CT or MRI. And as a reflection of these shortcomings and difficulties. Only a small percentage of patients with cirrhosis, probably only about 10-20% in the United States actually received twice annual imaging mm and I am frozen. Uh huh. All right. Good. Um So this data comes from a rare prospective study comparing the detection of HCC with ultrasound and MRI. And to make a long story short ultrasound only found 28 of the HCC's in this study that was conducted in Korea. Probably giving better results than we would have here in the United States because Koreans tend to be a little bit thinner so it gives ultrasound a little bit better of a chance. And because ultrasound is very, very widely used in Korea, um it is uh probably absolutely performing in this study. So we we were just to say, honey aroma swami Brooke quiet sarah lewis chip Bowman, who is one of the current residents in my PhD student, Nima did an investigation of participation in HCC surveillance in patients with advanced fibrosis and cirrhosis after they had been cured of hepatitis C with direct acting antiviral drugs. So the objective was to identify factors that were associated with HCC surveillance amongst erotic patients who have been cured of HIV HCV. And so in this study, we took imaging prior to cure and then eliminated people who had any suspicious lesions so that we could monitor participation in surveillance in a group of patients who had not already had an abnormality on their imaging study that might raise awareness and caused, okay, it's sort of an art if actual elevation of the screening rate. This was a higher preapproved observational study that enrolled adults with advanced disease, which we defined as a 54 score over 3.25 and clinical evidence of stage three or stage four fibrosis, who started treatment here at my hepatitis C treatment with direct acting antiviral drugs Between 2011-2018. And all the patients that we followed were cured of hepatitis C. So we were monitoring post your patients for incident liver cancer and participation in liver cancer screening. Uh Sarah Lewis reviewed all the abdominal images and as I mentioned, patients who had a suspicious ecclesia were excluded from the investigation of surveillance. And this is a picture of Sarah with two imaging uh images showing the washing wash out that is characteristic of liver cancer. So, uh we had a surveillance cohort of about 300 patients, 356 patients that were priests green. So they had no suspicious lesions prior to the time that they were cured of hepatitis C. And we then followed them For about three years. Um What we found was that patients without private insurance were less likely to complete two or three surveillance test um and that is shown here ah on yeah, you know, very analysis. Um the Failure to complete two or 3 tests uh was being african american or hispanic was a risk factor for failing to complete test. Um patients with higher platelets were less likely to complete two or 3 tests on multi variable analysis. African american race fell out, but private insurance persisted As a factor that was associated with failure to complete two or 3 surveillance tests. So at the end of the day, there are a series of hurdles that need to be clear to increase HCC early detection and reduce deaths through surveillance. So we're very focused on defining HCC risk factors more precisely. And we would certainly like to reduce economic and logistic barriers. Um, the radiologists here at Mount Sinai be sure to julie and Sarah Lewis and others are working on improved detection methods with an abbreviated MRI that should cut down the cost and speed things and uh I think we need to have a an awareness campaign to re examining re examine the surveillance guidelines. So I just want to close with, thanks to paul Burke and Barbara Murphy who passed recently, but we're certainly inspirational to me and I just thank them for their dedication and their visionary leadership and I'm happy to take any questions. Thank you. Dr branch, you think we already have a question in the chat? Can you read it to me? Let's see. I, yeah. Um from Dr Greenberg, has anyone looked for circulating tumor cells in black versus white or in any subgroups? I don't know. It's an interesting question. Certainly one can detect circulating tumor cells and as to whether the numbers differ, I just don't know. I don't think I've seen that analysis. I really don't know that it looked at that way. Two. Good question. Thank you. Other questions from the group comments. Is there another question? Andrew is there a role for diabetes and obesity and liver cancer? Oh, there definitely is. Um diabetes certainly increases the risk of lower cancer and in many cases there is less, like somewhat less likelihood in patients who have diabetes as a risk factor that the individual will have cirrhosis. Interestingly, diabetes appears to be less of a liver cancer Risk factor in african americans than it is in caucasians. But it definitely is a risk factor. Um specifically in patients with hepatitis C is the underlying liver disease, diabetes is less of a notable risk factor and that may be because hepatitis C itself is such an overriding driver of liver injury that the incremental increased risk that comes from diabetes get swamped out and it can't really be observed. But for sure in caucasians and to some extent in Hispanics, diabetes is an important risk factor for liver cancer and somewhat less so in african americans. Um Dr Greenberg gives a file common, which I'll let you know is one would think that micro vascular invasion would lead to circulating tumor cells and this comment and then we have another question, Dr Kohler says Andrey, great work. Are there any genetic differences in blacks related to genes in the fibrosis pathway? Yeah, I would so love to know that the answer to that question. Um, Not that I now that have been reporting, um it certainly seems very likely that there would be because there seems to be such a dramatic attenuation of the fiber optic response. That it's hard to imagine that it doesn't trace back to a germline difference. And um, I would very much like to investigate that. I think it would be helpful not only to understand liver cancer risk but just to understand the liver biology much better. So I'm really hopeful that those experiments can get done. Yeah. Dr Sachs says you mentioned insurance as a possible factor in this racial differences. Do you have any thoughts on other possibilities? I'm not sure. I understand the question. David, could you could you just reread the question please? Sure. Um, you mentioned insurance as possible factor in the racial differences. And then the question is, do you have any thoughts as to other possibilities to these different I would assume to the difference. Well, yes, I think that suspicion is uh, it's not easy to advocate to your patients that they should come in for twice annual imaging. And if you think that the risk is less because that patient doesn't fit your image of what a high risk person looks like. I think you you're urging of that patient to come in for imaging maybe less. And that's why I think it's so important for people to understand that this patient may not look to you as a provider like somebody who is at high risk because they don't have, they don't have those features that we associate with very advanced liver disease, such as societies evidence of portal hypertension. And so um in addition to the barriers that come from lack of insurance, I think there could be a barrier that comes from just lack of that. Your risk is higher than you than the provider and the patient imagines. Mhm. I'm Doctor Pickle has a question around what about the risk of insulin resistance in HCC. And then further goes on to say racial differences and insulin resistance question. Well, certainly they're anthro geometric differences between african americans, caucasians and Hispanics. And african americans are much less prone to central obesity. And generally speaking it is central obesity that is a risk factor for that pathway of insulin resistance that leads to liver cancer. So that overall body composition difference between african americans and other races I think is part of why the profile looks as different as it does. Um there's a particular genetic polymorphism in a gene P. N. P. L. A. Three which is associated with an elevated risk of liver cancer. And that genetic polymorphism is much more prevalent. And Hispanics, caucasians have kind of an intermediate prevalence and the prevalence is very low in african americans. So there are some genetic differences that are known to impact body composition and lipid trafficking in the liver. Um but I don't think they are all there is to this story. I think those genes identify a high risk group of Hispanics and caucasians, but they don't really explain the high risk in african americans who are at high risk. It's almost like a non overlapping set of risk factors that has not gotten delineated by race. Great. Any other comments or questions well. Andrew. I think that's all the questions and comments. Thank you very much for your very interesting talk today, letting us see all of your great research. Really appreciate it. Well, I appreciate the invitation and all the comments and questions. And if people have more comments and questions, I I'd love to hear from you, right, we'll have a nice day. Everyone enjoy the day. Take it easy. Nice seeing you all.