Good morning. It is my pleasure, together with my co-chair, Amit Segal, to present this session that is really devoted to the therapy in hepatocytic carcinoma. So, it's our pleasure to present the different presentations that we will have that are all recorded. First, I will speak on the molecular rationale for combination therapies, and we will have recharfine that will present a currently approved treatment for hepatocytic carcinoma, and then emerging therapy in phase 2 and phase 3 clinical trial will be presented by Joseph Leovet, and finally, Laura Kulik will present the management of adverse events in systemic therapy for hepatocytic carcinoma. So, remember that all the questions from the audience are more than welcome, and you can ask a question in the chat, and the question will be treated at the end of the session by online only. So, please ask a question on the chat. Thank you very much. Have a good session. So, good morning. It's a pleasure to try to ask a question. Do we have some molecular rationale for combination therapy in hepatocytic carcinoma? So, just to start with the background, I think to answer the question, we need to better know what is the mechanism of tumorigenesis in hepatocytic carcinoma, so in the liver. So, remember that we are accumulating mutations during life in our cells, and at the end, each HCC tumor is a unique, the result of a unique combination of genes that are altered, and that at the end, we have more than 40 different genes that are functionally altered in each hepatocytic carcinoma. So, now how we can use the knowledge of molecular classes and molecular classification into therapeutic targets, and why we should have to combine drugs. So, the agenda will be to review with you what are the molecular classifications of hepatocytic carcinoma, what we identified in advanced hepatocytic carcinoma, and what could be the result of the better knowledge of molecular defects in terms of new targets and tumor resistance. So, just to remind you that during aging, we have a progressive decrease of the telomeres in each cell of our body, and recently, we showed that by analyzing more than 1,400 non-tumor liver tissue from patients, we identified a close relation between the lengths of the telomeres that are these repetitive sequences that are found at the extremity of each chromosome, according to the age of the patient, and that the length of the telomere is more pronounced in males than in females, maybe accounting for the relative protection of the female from the risk of HEC, but it is more pronounced when you have short telomeres. The telomeres are shorter according to the severity of the chronic liver disease, and also according to the presence of alcohol consumption in the patients, and by doing multivariate analysis, the four determinants are independent in the control and independently associated with the length of the telomere. So, why speaking on that? It is because controlling the length of the telomere is very important to control the melanin transformation in hepatocytic carcinoma, and we know that telomerase is not activated in mature hepatocytes, whereas it is activated in 90% of the hepatocytic carcinoma, but what we know also is that during the natural history of the development of the chronic liver disease, telomerase activation through third promoter mutation occurs very early during this natural history, and that 6% of the low-grade dysplastic nodule in cirrhosis are already mutated for telomerase promoter, activating the expression of telomerase, and it's increased to 18% of the high-grade dysplastic nodule, and to more than 90% of the progressed HCC, and that the activation of oncogene and tumor suppressor gene occur later to ensure the diversity of, and the molecular diversity of hepatocytic carcinoma. So, really, third promoter mutation is the earliest recurrent alteration that we can find during the natural history of the development of hepatocarcinogenesis. Now, if we look at the molecular classification of HCC, all the molecular classifications that we described up to now with different publications, different teams, and laboratories across the world are related together. In my lab, we like this classification in six groups with more proliferative subgroups of tumor, G1 to G3, that are enriched in HBV-related tumor in TP53 mutation, and whereas G5 and G6 are less proliferative and usually associated with, in most of the cases, beta-catenin mutations through the CTNNB1 mutation activating beta-catenin. So, what is important to understand is that these molecular subgroups are really defining some homogeneous subgroups of tumors that are closely related to both clinical features, genetic alterations, but also to pathological features, so the phenotype of the tumor are also different according to the molecular subtype, and to biological pathway. And what is interesting is that these pathways could be targeted by specific therapy. This is the example of the IGF2 that encodes insulin growth factor type 2 that is overexpressed in this subgroup of tumors, and we know now that we have some inhibitors targeting specifically this oncogene, but we have also, for example, some beta-catenin pathways that is activated in G5, G6 that could also be the basis of specific therapy for the future. This molecular classification is also related to the prognosis of the patients and the intrinsic prognosis of the tumor, and the G3 subgroup of tumors is associated with prognosis, and we showed that it was also translated in pathology with Julien Calderaro that identified that the macrotrabecular massive subtype of histological subtype of epithelial carcinoma was in fact corresponding to the G3 subgroup of tumors, and it was associated with a particularly poor prognosis and poor survival in patients treated with liver resection or patients treated with percutaneous radiofrequency. So, it can really modify also how we can take care of the patients, and then the question for the treatment of advanced hepatocellular carcinoma was to have a better idea of the molecular classification into the advanced tumors and to compare this molecular classification with the less advanced patients. So, we performed this work with Georges Arnaud in the lab, and we compared the very advanced HEC with the less advanced one. So, what we identified is that a third promoter mutation or CTNV1 mutation were with a similar frequency observed with a similar frequency whatever the stage of the disease, but T53 mutations were more frequent in advanced hepatocellular carcinoma, and once again, in advanced hepatocellular carcinoma, we observed an increased number of G3 transcriptomic and molecular subtype of hepatocellular carcinoma, and it was related to the poor prognosis, interesting poor prognosis of this tumor. The surprise was to not using whole exome sequencing, we identified a similar number of mutations per tumor in advanced hepatocellular carcinoma compared to less advanced hepatocellular carcinoma. So, and it was also exactly the same for the number of chromosome arm aberrations. So, in other terms, there is no more, not more mutations, not more chromosomal alteration in more advanced hepatocellular carcinoma, but in fact, compared to the less advanced, but the type of mutation is different and because in advanced hepatocellular carcinoma, we identified frequent mutation in SF3B1 gene, in retinoblastoma RB1 gene, but also in T53 compared to the less advanced hepatocellular carcinoma. But what is also important to notice is that we didn't identify really a specific gene mutation that was found only in advanced HCC. Most of, all the mutations that we identified in advanced HCC were in fact also present in early hepatocellular carcinoma. But this game is really very important. You need also, I need to highlight that each tumor is really the result of different combination and different association of gene mutation together. So, it's not only different gene mutation, but for sure it is also different pathways that are altered together in the same tumor, but usually with a trunk mutation from early HCC to advanced one. So, interestingly also, what is important is to try to identify some oncogenic addictions. That means if we can identify some oncogenic alteration that are really very important for the survival of the tumor cells, because if the tumor cells are addicted to the oncogene activation, then we can hope to fight this oncogene addiction and to kill the tumor cells. So, it was a case for telomerase activation, for example. We treated cell lines with artisan oligonucleotide specifically directed against telomerase expression, and we showed in different cell lines that we were able to shut down the expression of telomerase and to decrease the telomerase and then to kill the cells. It was reproduced in all the cells that we have tested in more than 12 different cell lines, and we showed that it was really killing the cells with the activation of apoptosis, but also activation of the DNA damage. So, it is like telomerase is really an oncogene that is very important, very early during the beginning of the transformation of the hepatocyte, but also it seems that the transformed hepatocytes are dependent of the activation of telomerase. We can also look and search for different targets that we can hope to use for a therapeutic purpose. So, we have identified more than hundreds of genes that are altered in hepatocytic carcinoma, and among them, several of them can be targeted using specific drugs. This is all the genes with the yellow flags that can be targeted by specific inhibitors, but in fact, these inhibitors are targeting genes that are only mutated in less than 5% of the cases. So, it's very difficult to have a specific targeted therapy to be efficient in most of the hepatocytic carcinoma. So, the idea is really to, but if we put together all the gene defects, we can evaluate to 20% to 30% of the advanced hepatocytic carcinoma could be targetable by at least one of… because showing a genetic defect that could be targetable by an inhibitor that is already developed for and approved for other type of tumors. Interestingly, these targets are mostly targeting and activating the IKTM-TOR and the MET pathway, but also VAGF and FGF. So, this is just an example of how we can hope to use these targets in the future. So, it was a woman with HCV-related cirrhosis treated with sorafenid and RFA, with experience of multiple metastasis, and she was included in a clinical trial testing for tepotinib, and tepotinib is a specific MET inhibitor, and she experienced a progression… complete response during 14 months, and we showed that in the tumors, there were focal amplification of MET, and it was really because of the amplification and the high expression of MET that can explain the response to the inhibitor and to tepotinib. But the question is that after 14 months, the patient has relapsed, and so now the question is how to combat resistance to targeted treatment. Another example is coming from the FGF-19 pathway. FGF-19 is amplified in around 5% of the advanced hepatocellular carcinoma, and several companies have developed FGF receptor type 4 inhibitors, and this is the example of the physocatinib, and you see that a lot of response and partial response and complete response has been identified in patients that were positive for FGF overexpression, and it was not the case in patients that were not positive…for which the tumor was not overexpressing the FGF. But at the end, another paper in the same issue showed that all the patients in fact relapsed, and this relapse and the resistance to the treatment was in fact due to mutations that occur in FGF-4 gene at two major positions that confer resistance to the drug. So, once again, the question is how to combat resistance to targeted therapy. So, we have developed some screenings using cell lines, and we collected more than 30 different hepatocellular cell lines that were characterized by genomics, and we tested a large number of drugs, and we analyzed the results to identify some predictive markers of response to specific drugs. The idea was that, in fact, first by classifying the tumor, we found similar molecular classification that was observed in the primary tumors. So, in fact, cell lines are similar to the primary hepatocellular carcinoma cell lines tumors. We identified three different subtypes of cell lines that were unreached in TB53 mutation and FGF19 amplification. One of the limitations is that we have no good model of beta-catenin-activated tumor into the cell lines, and it remains to be developed for the future. But by doing this work, we can really analyze what are the sensitivities that are found into the cell lines according with specific drugs. So, this is an example of the panel of cell lines here that were tested for trametinib, that pinpoint that is an inhibitor of MEK that is downstream of the FGF amplification and signaling pathway. So, what we observed is that, in fact, the cell lines that were activated for FGF were, in fact, mostly sensitive to trametinib, and it is very interesting to notice that trametinib can bypass the resistance to FGF19 maybe in the future. So, this is prediction, but now also we need to validate the biomarkers that could predict the tumor resistance or tumor sensitivity to the drug. And by reviewing the literature, what we know about the tyrosine kinase inhibitors is that in the literature, several biomarkers and molecular defects have been found to be related to resistance or to sensitivity to the drug, the various drugs. But in fact, none of these biomarkers have been validated in clinical trial. Using immunotherapy, there is a lot of very good candidates. For example, the immune classification of hepatocytic carcinoma, there is only few macrosatellite instability, but it seems also that the number of mutations are not related to immune response. Does beta-catenin mutation predict the resistance to immunotherapy? This is really questions that remain to be validated once again in clinical trial, like other features and clinical features, such as NASH etiology of the HECs that remain to be validated in clinical trial. The only biomarkers that has been identified to be related to tumor sensitivity is in fact the alpha-fetoprotein, the high alpha-fetoprotein in the serum that is associated with the sensitivity to ramusirumab. That's very interesting because we can now select the patients with high alpha-fetoprotein to treat the patients with this drug. In conclusion, why combining drugs? First, because several pathways and genes are altered in each tumor HEC. In fact, each HEC is a result of a unique combination of altered genes. That means that we probably need to fight different pathways and to combine, for sure, targeted therapies that have limited results in terms of improvement of the survival with immunotherapies that demonstrate already, in terms of association, an increased efficacy for the treatment of the patient. Now, genomics really can help to identify molecular therapeutic targets that could be very useful, even if it is in a limited number of cases. Remember the story of MET, for example. But we need also to prevent resistant treatment. And one of the things that is very frequent in targeted therapy, but one of the best prevention of resistant treatment is also to start combined therapy. And it was already done in several other types of tumors by fighting different pathways that are associated together in the same tumor. So to finish, I want to thank my team and all the collaborators. Thank you very much. Hi there. I'm Dr. Richard Finn from the Geffen School of Medicine at UCLA. Thank you very much to AASLD and the organizers of this session to have me part of this distinguished panel. And I'll be talking about currently approved drugs in the treatment of HCC. Here are my disclosures. So we'll talk about the frontline setting first. Here you see the data with serafinib. Serafinib was the first drug approved in liver cancer back in 2008, where in two separate phase three studies, it demonstrated a significant improvement in overall survival versus placebo. Prior to 2008, no treatment had been shown to do this. And serafinib, a VEGF receptor multikinase inhibitor, an oral agent, was shown to improve survival by slowing progression. We did not have a really high objective response rate, single digit, but we learned that with molecular drugs, we could improve survival by slowing progression. And serafinib has a well-known side effect profile, which generally can be manageable, which is hand foot skin syndrome and in general GI toxicity. Now the first new drug approved in frontline occurred a few years ago, and that was lenvantinib, another multikinase inhibitor to the VEGF receptor that was looked at in this global phase three study in patients with advanced liver cancer, and randomized them to lenvantinib or serafinib. Now this study was powered for non-inferiority. And you see here that when we look at its primary endpoint of overall survival, lenvantinib had a survival of 13.6 months and serafinib 12.3 months. This did not meet its superiority endpoint, but did establish lenvantinib as non-inferior for the overall survival. The hazard ratio was 0.92, and the upper limit of the confidence interval was 1.06, less than the 1.08 cutoff for non-inferiority. So based on this data, the drug did receive approval. I should comment that lenvantinib did improve progression-free survival, and it had an objective response rate of about 19% as compared to serafinib, 6.5%. Now lenvantinib provided another option or does provide another option for our patients, not only because of its response rate and effect on PFS, but also has a different side effect profile. You can see here that hypertension is more frequent in higher grade with lenvantinib, whereas hand-foot-skin syndrome is more frequent in a higher grade with serafinib. In addition, we see things like diarrhea, anorexia, and weight loss and fatigue with similar incidences between both drugs. Novolumab is a monoclonal antibody to the PD-1 receptor, and it's been demonstrated to have single-agent activity in liver cancer. And here in the second line study of single-agent novolumab, we saw a response rate of about 15%, and for those patients who responded, they had a very durable response. Now just to step back and look at where novolumab and drugs like it are acting, and it's important to appreciate this because immunotherapies play such an important role in cancer now, including liver cancer. And you can see here, PD-1 ligand is expressed by the tumor cell, binds to the PD-1 receptor on CD8 T-cells, and tells these cells to essentially ignore the tumor. It down-regulates any immune response. And by interrupting this signal, we can then reverse this immunosuppression and restore anti-cancer effect. Similarly, there's other checkpoints, such as CTLA-4, and the CTLA-4 access is also being targeted by drugs such as ipilimumab and tremilumab. Now, because it had accelerated approval, we needed a phase three study to demonstrate in a randomized study that novolumab improved a survival outcome. The Checkmate-459 study looked at novolumab versus serafinib, and unfortunately, this was a negative study. Novo was not improved with novolumab, despite maintaining a response rate of about 15%. And you can see those results here. Now, recently, because the confirmatory phase three study was negative, the accelerated approval of novolumab was withdrawn, and single-agent novolumab is no longer FDA approved. But we did learn that the drug does have some single-agent activity, and this has become the benchmark for combination studies. In addition, the drug was very safe. We did see immune-related adverse events, but really at no increased frequency as in other diseases. And typically, these things are rash, hypothyroidism, hyperthyroidism, but really, any organ system can be involved. And all these patients are child PUA, and by doing so, really, we do not see any significant liver toxicity. Now, the big breakthrough in frontline treatment was the combination of atezolizumab and bevacizumab. This regimen received FDA approval in 2020 and has really set the new standard of care. And atezolizumab is a PD-L1 antibody, and bevacizumab is an antibody to the vascular endothelial growth factor. And by altering the immune microenvironment with a drug like bevacizumab, we see synergy with the anti-PD-L1 antibody atezolizumab. So the approval was based on the MBRAVE-150 study. This was a global study looking at atezolizumab and bevacizumab dosed intravenously every three weeks versus serafinib in an open-label design. The primary endpoints were overall survival and progression-free survival. Here you see the baseline characteristics. It was well-balanced. A majority of patients came from outside of Asia. All etiologies of liver disease were present. And because of concerns about bevacizumab and bleeding, because it hits the VEGF protein, patients were required to have an upper endoscopy within six months. If they had varices that were bleeding or at high risk of bleeding, they could not come on study. And you can see here about a quarter of patients in both arms had known varices at baseline and about 10, here you see 11, 14% of patients actually had them treated at baseline before coming on study. Here you see the primary endpoint. At the time of this analysis, follow-up was about eight and a half months. And we see a hazard ratio of 0.58 for overall survival. The survival curve separated early and remained separated. Now we updated this data this year now with over 15 months of follow-up. And we have the median survival in the atezobev arm, and it is 19.2 months as compared to sorafenib of 13.4 months. So a significant improvement in overall survival, and 19.2 months is now the benchmark for further drug development in the frontline setting. And we also improved progression-free survival as seen here on the lower right, 4.3 months to 6.9 months, a hazard ratio of 0.65. And this regimen, I would argue, is highly active. The objective response rate with this regimen is 30%, and that includes 8% of patients who actually have complete responses, meaning their tumor completely disappears. In addition, the median duration of response is over 18 months, really a striking data set that supports its use in the frontline setting. And I just want to show you the subgroups as far as objective response, some key things to notice. One, the response rate, regardless of etiology, was very similar, 13% with Hep B, 30% with Hep C, and 27% with non-viral etiologies. In addition, patients who had Barcelona stage B or intermediate liver cancer, so these would be patients who did not meet criteria for local regional treatment or progressed on local regional treatment, the response rate was 44%, quite high. Now, the regimen is well-tolerated. You can see here, as far as all adverse events, really not that different from serafinib. Grade 3-4 events, similar to serafinib. Grade 5 events, a little higher with serafinib. There were slightly more serious adverse events, 38% versus about 31% with serafinib. However, if you look at adverse events leading to withdrawal of serafinib, it was 10%. And if you had to stop both Atezo and Bev, it was only 7%. Here you see the most common side effects. You can see things that a patient would notice more, occurring more frequently with serafinib, such as diarrhea and hand-foot skin reaction, whereas hypertension and proteinuria were more common with Atezo and Bev. Bevacizumab as a VEGF agent is well-known to cause these side effects. And here, looking at all treatment-related adverse events that occur at more than 10%, again, you can see here, most frequently hypertension and proteinuria with Atezo-Bev, whereas diarrhea and hand-foot skin syndrome, rash, anorexia, are all more common with serafinib. And that supports this improvement in patient-reported outcomes with Atezo-Bev versus serafinib. You can see here that looking at time to deterioration quality of life, markedly improved or delayed with Atezo-Bev. Now, bleeding events are more common with Bevacizumab. However, most of these are low-grade, such as epistaxis. There were a few high-grade bleeding events, including upper GI events with Bevacizumab, but these were really single-digit and not very common. And again, the way to mitigate this is making sure patients have no high-risk varices at the time of study enrollment. And for those patients who did have high-risk bleeding events, the majority of them actually had main portal vein invasion. Now, we see that patients with main portal vein invasion do still get a survival benefit, perhaps. These are the ones that should be watched a little more closely. So in second line, we have many drugs available. The first drug to be approved in second line for patients who progress on serafinib was regorafinib, another multi-kinase VEGF receptor inhibitor, which was compared to placebo. And it was shown to improve survival versus placebo in patients who had documented progression on serafinib and tolerated serafinib for a minimum period of time. Here you see overall survival was improved from about eight months to 10.5 months with a hazard ratio of 0.62. And like other kinase inhibitors, it did this without inducing an objective response rate of great significance. Here, only 7% of patients. The side effect profile with regorafinib is consistent with other drugs in its class, hand-foot-skin syndrome, hypertension, fatigue, and also some GI toxicity. Cabizantinib is a VEGF receptor inhibitor, as well as an inhibitor of two important kinases in the immune microenvironment, such as Axil and Cmet. And this also was compared to placebo in a second line setting. This study improved overall survival with a hazard ratio of 0.76. And you can see here an improvement from eight months to 10.2 months. Unlike other second line studies, this study actually had about 25% of patients who would even be considered third line. And again, it improved survival mostly by delaying progression as the objective response rate here is low, only 4%. And the side effect profile with this kinase inhibitor is exactly the same as we've seen with other kinase inhibitors listed here. Now the other drug approved on phase three data is ramicirumab. This is a monoclonal antibody to the VEGF receptor. Ramicirumab was initially studied in a second line setting in all comers. And that was a negative study, the REACH study. However, retrospectively, it was noticed that patients who had an elevated AFP might get a benefit from this drug. You can see here patients who have a high AFP in the placebo group lived only 4.2 months, whereas those with a low AFP lived 12 months. And it was in this high AFP group that we saw an improvement in overall survival. Now this data supports the idea that high AFP is prognostic based on the placebo data, but it also suggested that perhaps high AFP is predictive of response to ramicirumab. So that led to the REACH-2 study, which was a second line study after sarafnib and randomized patients to ramicirumab intravenously every two weeks versus placebo in a blinded fashion. This study met its survival endpoint with a hazard ratio of 0.71, which is a 29% decrease in the risk of death. The absolute improvement median was 7.3 months versus 8.5 months. The control arm here was somewhat small, only 95 patients. And you can see here that the placebo group of 7.3 months behaved better than we expected based on the retrospective analysis. However, ramicirumab provided the same survival, if not a little better, at 8.5 months. Now ramicirumab is very well tolerated. It is not a kinase inhibitor, it's a monoclonal antibody. So all of its side effects are really related to its on-target effects. It can cause fatigue, hypertension, and fluid retention, which we can see as edema, and some increase in ascites. But things like hand-foot skin reaction or diarrhea do not occur with great frequency. And if they do, they're low grade, at least the diarrhea. Hand-foot skin syndrome we do not see. Now pemoralizumab is a PD-1 antibody like nivolumab. It received accelerated approval based on this study with an objective response rate of about 18%. For full approval, a randomized confirmatory study had to be done, and that's the Keynote 240 study. This was pemoralizumab versus placebo. This study confirmed the objective response rate of 18% and the long duration of response of here over 13 months. Here's the primary endpoint of overall survival. And we see here an improvement of 10.6 months in the placebo arm to just under 14 months in the treatment arm for a hazard ratio of 0.78. The upper limit of the constant interval was 0.998, and the p-value was 0.02338. However, to be declared positive, this study had to have a p-value of 0.0174, and it did not reach that. There were several interim analyses as well as a split alpha between overall survival and other endpoints that required this low p-value to be considered positive. But clearly, there was a very strong trend. The ODAC, Oncology Drug Advisory Committee, reviewed this data and did recommend that pemoralizumab maintain its accelerated approval in second line pending confirmatory. See a press release for the confirmatory study, Keynote 394. This is one of the confirmatory studies, which was pemoralizumab versus placebo. And this was done in an Asian population. And this study met its primary endpoint of improving overall survival and other endpoints such as progression-free survival and objective response rate. And so this supports the activity of pemoralizumab in second line. It's actually the first level one evidence for single agent PD-1 inhibition in advanced liver cancer. And we look forward to seeing this data. Now, the other drug regimen that's approved in the second line setting is nivolumab, the PD-1 antibody, in combination with the CTLA-4 antibody, ipilimumab. So this is two intravenous immune checkpoint inhibitors. And in this single arm study, looking at three dosing, single treatment, I should say, but looking at three dosing arms, all of them demonstrated a response rate of about 30% that was quite durable. And on the survival curve in the second line setting, you can see in arm A that the survival second line was about 23 months. This is very impressive and was used to support the accelerated approval. The confirmatory phase three study is ongoing, which is nivo-ipi versus TKIs in the front line setting. And here you see the treatment-related adverse events. We see more side effects with the combination of nivo-ipi versus single-agent pembrolizumab or single-agent nivolumab. In those single-agent PD-1 studies, those drugs tend to be very well tolerated. Here you can see an increased frequency of various toxicities, including diarrhea, enzyme increases, rash, pruritus, fatigue, and others. If we concentrate on the immune-related adverse events, you see them here. And actually, in this arm A or across the study, I should say about 50% of patients do need steroids to manage an immune-related adverse event. These are manageable and patients do recover. And at that point, they could be considered to be re-challenged, but I think Laura Kolick will go into adverse event management in her talk. So in closing, we have several regimens available. We've improved survival in front line, and for those patients who can't tolerate atezobev, we offer soraftib and linvantib as options. But ultimately, I think we'll improve survival by sequencing patients through all the active drugs we have available, as long as they can tolerate them. So it's important for patients who present with intermediate disease and get local regional treatment that we transition them to systemic treatment at the right time. We cannot do local regional treatment beyond progression and compromise their liver function, which then might limit them receiving drugs that are active and improving survival for advanced disease. So in conclusion, we've made a lot of progress in the advancement of liver cancer, really, especially in the past few years. Really, we need to consider atezobev, and unless there is a contraindication, that should be the treatment of choice. But if there is, linvantib and soraftib are good options. We have several regimens approved in second line, but in reality, will probably become second, third, and beyond. And these include Rego, Raffinib, Cabizantinib, Ramicerumab. Pemrolizumab is approved currently on accelerated approval as is Ipilimumab and Nivolumab. And how to optimally sequence these drugs really is not known and probably real world datasets will tell us how to do that. But really, again, I think sequential active agents will ultimately continue to improve survival. And it's important to start these treatments at the right time. So I want to thank you again for your attention. I hope everybody is well. Hopefully we'll be in person very soon. And I want to acknowledge all my colleagues at UCLA who have helped me in our research efforts in helping to improve outcomes for all our patients with liver cancer. Thank you very much for your attention. Hi, good morning. I am Professor Llobet, I'm Professor of Medicine and Director of the Liver Cancer Program at Mount Sinai, Professor of Medicine, University of Barcelona. I'm going to talk today about emerging therapies in phase two and phase three in advanced HCC. We'd like to thank the organizers to invite me to give this lecture. Here we have the outline of my presentation. First, I will talk about first and second line standard of care. As you know, since the publication of the New England Journal of Medicine paper on sorafenib in 2008, sorafenib showed survival benefits compared to placebo, 10.7 versus 7.9 months with a hazard rate of 0.69 and become the standard of care in advanced HCC. Later on, several trials have been reported yet I'm showing the list of trials in front line and in second line. In green, you have the positive trials. In orange, you have the positive trials for non-inferiority. And in red, the negative trials. As you see, beyond sorafenib versus placebo, in front line, we have lemba versus sorafenib with a non-inferior design that hits the end point. And very recently, atezolizumab plus bevacizumab showing superiority to sorafenib. In second line, only three trials have shown superiority, cabozantinib versus placebo, regorafenib versus placebo, and ramosinumab versus placebo only in patients with biomarker AFP more than 400 nanograms milliliter. The mechanism of action of all these drugs are summarized in this cartoon. As you can see, we have the neoplastic cell here in the middle, T-cell on the left, and then the allele cell on the right. Atezolizumab, atezolizumab is blocking PD-L1, bevacizumab is a monoclonal antibody against the ligand, BHGFA. Ramosinumab is blocking the receptor 2 of BHGF. And then we have the tyrosine kinase inhibitors, sorafenib is a multikinase inhibitor blocking platelet-derived BHGF and RAS. Lemba also is blocking FGF receptor, particularly receptor 4, which is important in the pathogenesis of HCC. Regorafenib has a similar profile, including, as you can see here on the right, in the endothelial cell blocking type 2. And finally, cabozantinib is blocking that. We also have that sorafenib and rego are blocking RAS signaling. On the left, you have the drugs approved by FDA based on phase 2 data. Nivo and Pembro approved Nivo in combination with ipilimumab that is a CTLA-4 inhibitor. I'm going to talk about first the drugs that have been approved. The mechanism of action of these drugs, particularly now I'm going to talk about atezolizumab plus bevacizumab is that we have around 70% of HCCs are cold tumors with very marginal infiltrate of T-cells, and they are barely responding to checkpoint inhibitors as opposed to 20% to 30% of the tumors that are hot tumors. How we can switch a tumor from cold to hot tumor? So what we can do is combine checkpoint inhibitor with a VHF inhibitor or tyrosine kinase inhibitor of the kind that are approved in HCC, sorafenib, rego, lemba, or cabo. And some of these combinations are able to A, decrease Treg activity, increase cytotoxic activity of T-cells, increase proliferation maturation of dendritic cells, decrease M2 phenotype macrophages. And with these, these drugs may be able to enhance activity of checkpoint inhibitors and switch a cold tumor to a hot tumor. And this is what it is known that happened with this trial. Tesolizumab plus bevacizumab, the single agent activity was modest, but in combination, they lead to superior benefits compared to sorafenib in terms of overall survival. As you can see, 19.2 months for the tesobep, 13.2 for sorafenib, which has a rate of 0.66. Also, the patient-reported outcome on the progression-free survival endpoints were hit. In this case, I'm showing here time to quality-of-life deterioration, 11.2 months for tesobep, 3.6 months for sorafenib. Here you have the summary of all the drugs in the front line and in the second line, it has the ratios in green, the trials that have been positive, and in orange, the trials that have been positive for non-inferiority. And finally, the current strategy for treatment in patients with untreated viruses or any contraindication for immunotherapy, eventually, sorafenib and limbat are still front line, but if they do not have these contraindications, which will be the majority of patients, it is estimated around 80 percent of the patients will be treated in front line with tesobep and follow up on progression with sorafenib or limbatinib. Conversely, there are other guidelines that my recommend not only consider sorafenib or limbat, but also Rego, Cabo, and Ramucirumab, and in the United States, Nivo, Ipi, and Pembro are also approved in second line. Let's talk about now new trials and emerging therapies. First, this is the treatment strategy for HCC and the staging system according to BCLC, zero, A, B, C, and D, and below you have the trials in adjuvant mode in the intermediate stage and in advanced stage. As you can see, all these trials, around 20 trials, might be able to change the field and become new standard of care, and it will be certainly the first time for an adjuvant therapy or even for a therapy in combination with this at intermediate stage. I want to specifically focus on advanced HCC. Here we have the trials in first line that we have at least seven trials, and also in second line, patients progressing a tesobep are currently already several trials ongoing in this set. I'm going to talk about some of these combinations. For instance, the first one, Limbatinib-Pembrolizumab, and also Nivol-EP and other combinations. In terms of Limba plus Pembro, this is the GCO. We published in 2020 100 patients treated with this combination, and interestingly enough, one of the key results where the objective response rate were around 40, 45 percent according to investigator and central radiological review, and also disease control rate was almost 90 percent. These are remarkable results that were followed by a progression-free survival of 9.3 months and overall survival of 22 months. This provide the rationale to move to the LEAP-002 study that's currently ongoing, randomizing 750 patients to receive this combination compared to Limbatinib alone as the standard dose, 8 milligrams per day, in patients with weight of less than 60 or 12 milligrams per day, and the certification factors are listed here. The other combination is Nivolumab plus CpLimumab. As you can see here, three treatment arms were tested. In terms of objective response, around 30, 32 percent of objective response was seen, and the duration of response was, in all circumstances, beyond 15 months. What is more interesting is that in the first regime, the Nivolumab EP3, the median survival was 22 months of median survival, and also in terms of overall survival according to the response. What we can see is that patients achieving objective response, this 30, around 32 percent of patients achieving response, they have a median survival beyond 33 months, which is a remarkable signal of efficacy. So, as a result of that, the checkmate A9DW is currently ongoing comparing Nivolumab versus Sorafenib or Limbatinib. Another combination that have been tested in phase three is Cabozantinib plus Atezolizumab, Cabozantinib at a dose of 40 milligrams orally, slightly lower than the standard in second line, and versus Sorafenib, and also there is an arm of Cabozantinib at 60 milligrams orally. We know for a recent press release of this phase three COSMIC-312 trial that the interim analysis show positive results with significant differences for progression-free survival with a hazard rate of 0.63 and a p-value of 0.001, but no differences, statistical differences, were seen in terms of overall survival, and therefore the final analysis is currently waiting for being read. Another alternative is the, and I'm showing here the data of the phase one two trial of Durvulumab versus Tremelinumab, a different dose. Interestingly enough, this dose of the TREME 300 was the one that moved to phase three because of this median survival in second line of 18.7-month median survival, and as a result of that, the Himalaya trial is currently still ongoing in terms of following the patients for events to happen with three arms. The Durva alone and Durva-TREME or sorafenib monotherapy. So, and as a result of that, the Himalaya trial is currently ongoing, comparing three arms, Durvulumab, TREME-Durva versus sorafenib monotherapy. Also, we have the Sintilimab plus Bevacizumab study. This study was reported very recently, and it compares this combination of a Chekovin inhibitor versus Bevacizumab, similar to the TESOVAB versus sorafenib in China with a co-primary endpoint of OS and progression-free survival, and the results have been released and actually reported, as you can see. This was a two-to-one randomization in Asia, and as you can see here, the progression-free survival was 4.6 months for the combination compared to 2.8 months for sorafenib alone, and the overall survival was also significantly different. It has a rate of .57, a P-value of less than 001, and actually, the median survival for sorafenib was 10.4 months but was not reached for the combination. Another trial that has been positive, reported also recently in GCO, is donafenib. It's also a multi-kinase inhibitor, particularly blocking BHG receptor 2, and it was compared head-to-head to sorafenib with a primary endpoint of overall survival that was hit, as you can see here, with a hazard ratio of 0.83, 12 months for donafenib, and 10.1 months for sorafenib. This leads, and this led to approval of this drug in China. Let's move to a single Asian checkpoint inhibitor. The Nivo sorafenib trial check made 459. We have the results in terms of objective response for Nivo, 15%, 7% for sorafenib, but when we were checking overall survival, as you can see here, the median survival for Nivo was 16.4 months, 14.7 months for sorafenib. Certainly, these outcomes are outstanding, and at that time point were the best outcomes reported for the phase 3 trial in single Asian in frontline, but the hazard ratio was 0.85, and the p-value 0.07, non-significant. In second line, something similar happened with pembro versus placebo. The objective response was 18% for pembrolizumab, but the overall survival did not hit the primary endpoint because the pre-specified value for a positive trial was 0.017, and the value of the P was 0.02, with hazard ratio 0.78. Nonetheless, the median survival of pembrolizumab in this trial was 13.9 months. That is remarkable considering we're talking about second line, 10.6 months for placebo, and very recently, another trial that was conducted in Asia, the Keynote 394, showed survival differences in this press release that was out in September 27. This trial confirmed benefit for pembrolizumab versus placebo in terms of overall survival, progression-free survival, and objective response. In terms of second line, we have the study comparing regorafenib plus pembrolizumab in second line, single arm. Remarkably, in second line, a partial response was achieved in 31% of the patients, and in terms of adverse events, or treatment-related adverse events, accounted for 41% of the cases grade three at the dose of regorafenib of 80 milligrams as opposed to 66% at the dose of 120, and leading to dose interruption here was at 23% of 37. So, therefore, the dose of 80 milligrams appears more appropriate for this type of approach. Another study has been the triplet, NEBO plus EP plus minus cabozantini, and here you have the outcome for the doublet and the triplet. The median survival for the doublet was 21.5 months. The triplet was not reached, and for progression-free was 5.4 and 6.8 months, respectively. In terms of objective response, was 19% assessed by investigator at the doublet, but 29% for the triplet arm. In terms of emerging treatments, of course, I'm listing here a lot of combinations or even single agents that are considered in combinations or with target therapies, checkpoint inhibitors with checkpoint inhibitors, triplet combinations, novel immunologic agents, and also other strategies, including CAR T cells or oncolytic viruses. I'm showing here the results of the study in phase two of carmelizumab plus apatinib that also lead to very good objective responses, as I'm showing here, and both in first line and in second line, those patients achieving complete or partial response that were around 30% of the patients have an outstanding outcome, as I said, both in front line and in second line, and the phase three study is currently ongoing. Also, I want to point out other approaches. For instance, the phase three focus trial in the formal pre-planned futility analysis for this study, this continuation was recommended. For this pexabec virus, the study has been concluded. Finally, other endpoints and other unmet needs are, first of all, we know now that for all these trials, second-line therapies are used in 50% of the patients, and we checked whether there was a correlation between a progression-free survival and overall survival in order to understand if this endpoint could be rescued for the design of trials in phase three. As you can see here, just analyzing 21 randomized controlled trials in the green area, there is extreme correlation with positive survival benefits. If the hazard ratio is below 0.6 for progression-free, this was reproduced with 27 trials, and we think that this will be the proposed scenario in green. If you have a hazard ratio at the end of being less than 0.6, it could be safety to say that correlates with survival. Above 0.7, it does not correlate with overall survival, but hazard ratio between 0.6 and 0.7, it is still uncertain. Finally, we want to talk about the results of checkpoint inhibitors in NASH-related HCC. As you know, NASH now is one of the theologies that is growing 20 percent in the West, and in one-third of the patients can induce HCC without the step of cirrhosis. In this meta-analysis that we conclude with 1,600 patients in three randomized controlled trials, what we saw is that in viral-related HCC, the overall meta-analysis show hazard ratio of 0.64 with an upper boundary of 84, whereas in the non-viral, the hazard ratio was 0.92 with an upper boundary of 1.1, and the P of interaction was 0.03. So, we can conclude from this that immunotherapies might actually respond significantly better in viral-related HCC as opposed to non-viral. This doesn't mean that doesn't work in non-viral. And it has been questioned whether it might be that this is not a predictor of response but actually a pronostic factor, and we have reproduced that in the three trials with immunotherapies and compared that with what happened in trials assessing TKIs, and certainly the median hazard ratio was similar for viral and non-viral when testing TKIs but was distinct when testing checkpoint inhibitors. And therefore, that is something particularly related to NASH HCC that may lead to less responses. So, in conclusion, atezoplasmabacizumab is the first-line therapy. Sorafenipalemba are used in case of contraindication or progression to the combo. And afterwards, regalcavonramone should be used. Embryo and NEBO-EP are FDA-approved. In terms of new combinations, we have the COSMIC 312 hit the progression of this viral endpoint, and the other trials I'm listing here are still ongoing. There is a trial combining a checkpoint inhibitor, Sintelimab plus Bevacizumab that was positive compared to sorafenib, and also donafenib versus sorafenib, and this has been approved in China. And then Pendrolizumab in second line that is FDA approved based on phase two. Now the trial that was negative in the West has been positive in China. In terms of what med needs, emerging therapies in combination are tested in all stages. We should consider progression-free as new primary endpoint in front line, and there is a need for new approaches in patients with NASH HCC. With this, I'm finishing here. Thank you very much for your attention. Hello, my name is Laura Kulik, and I will be talking about the managers of adverse events and systemic therapies in HCC. I will specifically be focusing on adverse events related to immunotherapy. I am a transplant hepatologist at Northwestern. I focus on the care of patients with HCC, specifically trying to get them to downstage or bridge to transplant. When I'm not working, my hobbies include hanging out with my two Frenchies, Miles and Davis. These are my disclosures. So I've listed the immune checkpoint inhibitors that have been used in HCC. When we compare CTL4 to PD-1 or PD-L1, we see that CTL4 has higher rates of toxicity, and the onset of toxicity is shorter, and specifically, it varies with the dose of the CTL4. Now, this contrasts to PD-1 or PD-L1 inhibitors, which is independent of dosing. Immune-related adverse events can affect any organ in the body, and it can occur at any time, although for each toxicity, there is a known average time period, but it's important to know that these toxicities can occur after discontinuation of immune checkpoint inhibitors. Fortunately, these can be treated and are rarely fatal. A meta-analysis of nearly 11,000 patients looked at patients who were treated with checkpoint inhibitors and compared to non-checkpoint inhibitors, and they found that there was a high grade of colitis and AST elevation associated with checkpoint inhibitors, which was statistically significant compared to those without checkpoint inhibitors. And fatal events were less than 1%, but in this meta-analysis, there were no HCC patients included. I'd like to draw your attention to three excellent resources on this topic. They all give their own algorithm in terms of how hepatotoxicity should be managed, which I will go through. First, I want to just briefly talk about the immuno-checkpoint inhibitors and why this leads to toxicity. When the breaks of the immune system are unleashed, specifically CTL4 and PD-L1, and they are no longer inhibiting the T cell, we then can have autoimmune events occur. The CTL4 is early. This is in the priming stage, and therefore, there's more widespread and severe events, whereas the PD-L1 is more specific and therefore causes less widespread events. These highlights the different organs that can be affected. As I said, any organ can be affected, and the pink asterisks are the ones that are mostly affected, the skin, the GI tract, the endocrine system, and the liver. But it's important to know when seeing a patient on these medications that virtually any organ can be affected by these medications. Now, we know that HCC predominantly occurs in patients with underlying cirrhosis, and in patients with cirrhosis, they are often plagued with significant symptoms that can overlap what we see with immune-related adverse events. So looking at these, I won't go through all these, but you can see that there are many different organs that are affected by cirrhosis and the related symptoms, and therefore, it's going to be very important to try to discern if this is indeed related to the immune checkpoint inhibitor or due to the cirrhosis itself. So this leads to, in making a decision, if one were to continue checkpoint inhibitors, there is the concern of worsening the immune-related AE, leading to delayed recognition, therefore, increased toxicity. And some of this delay comes from doing invasive procedures such as biopsies, colonoscopies, bronchoscopies, or waiting for culture results as opposed to starting steroids in a timely fashion, as opposed to discontinuing immune checkpoint inhibitors prematurely, which would lead to loss of efficacy and the side effects related to immunosuppression, specifically in patients with cirrhosis who are immunosuppressed by definition. And then comes up the question of re-challenging patients after they've had an immune-related event. So this is a very nice compilation of the different considerations when you're specifically looking at a patient with HCC, and it shows the immune checkpoint adverse events versus the side effects associated with chronic liver disease or the progression of the specific cancer or others. So it is really important to go through this exercise when trying to decide what the patient is actually experiencing. This highlights the most common immune checkpoint inhibitors, the skin, specifically rash, followed by the GI tract with colitis and diarrhea, the liver and hypothyroidism. And here on the bottom, we see the frequency based on what patients are receiving. In the purple, this is combined CTL-4 and PD-1. Here is CTL-4 alone. And in the orange, we have the PD-1 alone. Liver, you can see, is not the predominant. It is in the darker colors. And we can see that the highest one is, again, with the combined CTL-4 and PD-1. So this is a summary taken from the ASCO guidelines that were published in 2018 on the management of the immune-related AEs and the grade. And when someone has grade one, you can continue and follow them, watching them. If they are grade two, it is recommended to suspend the immunotherapy and start low-dose prednisone. If they become grade three, then transitioning to high-dose prednisone and discontinuing the immune checkpoint inhibitor permanently. However, if it is an endocrine immune checkpoint inhibitor event, then patients can continue, regardless of the grade of the toxicity, as long as there's stability with hormone replacement. And then if patients are still symptomatic after being on high-dose steroids, there is a reference to ASCO and an organ-specific guide as to what second-line therapies may be. I want to specifically point out that steroids should be tapered very slowly in order to make sure that the symptoms do not return. And prophylaxis against bacterial infection should be considered. It has been shown that 7% of patients on immune checkpoint inhibitors develop serous infections, and the most common cause was in patients who were on steroids or a form of immunosuppression. This just highlights the management of non-liver toxicities that I just kind of wanted to point out. I'm mostly going to focus on the liver, but in patients who have a mild toxicity, if this is pneumonitis or myocarditis, one should consider withholding the immune checkpoint inhibitor, even though it is very mild. Again, with grade two, if this is pneumonitis or myocarditis or peripheral neuromotor symptoms, one should consider permanently discontinuing, and all others permanently discontinuing when this is grade four. It tells how often this should be monitored, that patients should be referred to the specialist for that specific organ once they have a grade two, and talks about the dosing of steroids for each grade. Now, looking specifically at immunotoxicity from immune checkpoint inhibitors related to the liver, we have seen that these are higher in patients with HCC compared to non-HCC tumors, and this is maybe confounded by the fact that patients have underlying liver disease. It may be due to tumor progression within the liver, and there are fluctuations in patients' liver tests when they have underlying liver disease. The onset of this is generally four to 12 weeks after starting an immune checkpoint inhibitor, but it is important to realize that this may start late as well, and the pattern is very heterogeneous. It can be hepatocellular, cholestatic, or mixed, and it really depends on which immunocheckpoint inhibitor you are using. If it's a PD-1 or PD-L1, this tends to be an increase in the liver enzymes, and the onset is generally 14 weeks, and you can have a fever about 11% of the time. As opposed to CTL-4, there is more cholestasis. The onset is quicker at three weeks, and fever is more commonly present in about three-fourths of the patients, and when you look at pathology, which we really have a paucity of data on this because biopsies are often not done, it is a very nonspecific pattern in patients who are PD-1 inhibitors, whereas in CTL-4, this is granulomatous hepatitis and often with central vein endotheliitis, and when there is staining done for CD-8 and CD-4, CD-8 is greater than CD-4 in CTL-4, whereas there is no difference in the PD-1, PD-L1 inhibitors. Now, this is looking at the grading. NCI uses the typical grading that oncologists have been very used to. As hepatologists, we know that if the enzymes are elevated, the degree of enzyme does not predict what you will see on a biopsy, nor does it predict how the patient will do. We take into account what their bilirubin is as well as INR, and this has been used in DILI, looking at the different grades, taking into account patients' bilirubin and IR as well as development of symptoms such as encephalopathy and the duration of jaundice, being jaundice. The risk factors for checkpoint inhibitor toxicity, which was specifically seen in melanoma, were patients who had prior autoimmune diseases, but interestingly, this was mostly rheumatologic. There was no autoimmune hepatitis, and this is in patients who were treated with CTL-4 inhibitors. It was dose-related in CTL-4 inhibitors, and low doses of IL-6 in CTL-4 inhibitors was a risk factor for hepatotoxicity as well as combination of immune checkpoint inhibitors. So looking at the management of hepatitis in HCC, here are the numbers in terms of increase in AST that we're seeing based on monotherapy as well as combined therapy. Today, we just found out that Dura and TREMI, this was a positive study compared to sarafinavolone, so we'll see more of this being used in clinical care. This is a chart from Sanger and colleagues, and what it does is it recognizes that patients with liver disease will often have increased enzymes, and therefore, when you are looking at how the degree of the grade is, it takes into account that they had increased enzymes to start with. It also takes into account what their bilirubin and INR is and tells you what should be done in terms of steroid management. I find this to be a very helpful guide when seeing patients. We also need to look at differential diagnosis. What could be other causes for hepatotoxicity? We have patients with viral hepatitis, other hepatitis such as CMD, EBV, is there serendipitous drinking? Is there hepatitis related to ischemia? Is the bile duct blocked? Is there PVT? Has the tumor progressed? And then also looking, is the degree of AST to ALT elevation? There may be myocarditis or myositis that is causing the AST to be more elevated. Not the liver at all. Now we have a paucity of liver biopsies that are done. The pro is it can help rule out other causes that are shown above and confirm to some degree that this is related to the immune checkpoint inhibitor. But unfortunately, the histology is not pathognomonic and there can be a risk of charging steroids as you're waiting the biopsy as well as complications of the biopsy. But doing this could lead to advancement in knowledge in this area. This is another proposed algorithm. This comes from Charlton and colleagues. And it looks specifically, I wanted to highlight that they say to consider a liver biopsy at grade two, which is slightly different from what I'd shown you. They say to hold the immune checkpoint inhibitor, but no immunosuppression should be start. It has been shown that patients, when you stop immunotherapy, that they can improve without the addition of steroids. However, if they are not continuing to improve, they state to start low dose steroids. And if they continue to not improve, higher dose steroids. And they also take into account re-initiation of the immune checkpoint inhibitor after there has been improvement in the liver function. Lastly, this was from D. Martine and colleagues. And it's looking, starting at patients who had severe LFT derangements and taking into account that there was no other cause. And they also looked at what the bilirubin and INR was. And if they did not meet this, they would continue to follow the patient. So this is even a little bit further, being that it is severe. If it is not improving, starting with low dose steroids. If there is no improvement, starting with a higher dose steroids, then MMF, and then even adding tecrolimus. And on the side here, you can see the taper. The taper is a little bit longer for the more agents that you had to use in order to see improvement. And they do take into account when you may be able to restart an immune checkpoint inhibitor. However, if a patient had to go all the way to using tecrolimus, it is recommended to stop permanently. Now, what about the use of steroids on outcomes? There has been some data that this may have a negative impact. This was a meta-analysis of 16 studies looking at steroids versus no steroids in patients who were treated with immune checkpoint inhibitors for various malignancies. And the hazard ratios was 1.5, with a higher risk of worse outcomes in patients who receive steroids. And when they specifically looked at what was associated with this, in patients who were given steroids for supportive care to treat their symptoms, this was associated with a worse outcome. Patients who were given this for brain mats, again, associated with worse outcome. But patients who were given steroids to treat immune-related AEs, there was no significant association. And there has been thoughts that patients who have AE, similar to what we've seen in TKIs, that this is prognostic and is showing that the immune system is working and that these patients have a improved overall survival when they have immune-related events. Another study looked at patients who were treated with steroids, and they looked at the time from when they were started in conjunction when the immune checkpoint inhibitor was started. And what they found is that both overall survival and progression-free survival were worse when the steroids were started sooner. So if the steroids were started within two months of the immune checkpoint inhibitor, there was an overall worse survival and progression-free survival. So timing may be important for the immune system to get started. What about the safety after someone has had a grade two or higher? This is looking at 93 patients of which 40 were re-challenged. Of those 40, 22 patients had recurrence of their immune-related adverse event with a median of 14 months. You can see here that the majority of the immune-related events were the original event as opposed to a new event. And this kind of highlights this in terms of which were new, which were old, et cetera. What they found is that the median time from the initial event was significantly associated with higher risk of having a second event of nine versus 15 weeks. And importantly, they found if you do have a second event, that it was not more severe than the initial event. So there are some questions specifically about re-challenging patients with immune checkpoint inhibitors after they've experienced hepatotoxicity. Which agent should you use? There has been some evidence that there is a higher rate of recurrence of adverse events if the patient had a PD-1 inhibitor followed by a CTL-4 inhibitor. And this has been hypothesized that the PD-1 inhibition continues even after the half-life has deteriorated and there is no detectable PD-1 in the system. So by adding a CTL-4, this may be essentially putting patients on combination therapy. What is the appropriate time for normalization of LFTs until restarting? There may be some data that waiting a little bit longer. And is there a benefit of prophylactic immunosuppression? One study found that there are higher rates of toxicity in patients who were on combined steroids with immunocheck inhibitors as opposed to no steroids. And could there be a role of Budesonide, which may be controversial in patients who have cirrhosis because of the effect of first pass on this medication. And this looks at the question of autoimmune disease. Patients in the trials, none of them were allowed to have autoimmune disease. So in clinical practice, there has been articles that have been published regarding the use of immune checkpoint inhibitors in patients with baseline autoimmune disease. And what they found is that it's effective in patients with autoimmune disease compared to those without autoimmune disease or the general population. However, adverse events related to checkpoint inhibitors may be higher, but manageable, and that patients need to be followed closely. To this end, there has been a proposed two-step strategy that when you specifically look at what the actual autoimmune disease is, using the appropriate select immunosuppressant and then having an algorithm to optimize the immune checkpoint in that specific autoimmune disease here. And I will not belabor this point, but you can see that these have been really well thought out and it really highlights the need for a multidisciplinary approach. There is now a ongoing trial looking at patients with preexisting autoimmune disease who are receiving Nivolumab and patients with advanced cancer. You can see that these are patients with Crohn's, dermatomyositis, rheumatoid arthritis, multiple sclerosis. So a wide range of different types of autoimmune diseases where we will be able to see what the outcome is in treating these patients. And I wanted to highlight that the toxicity from immune checkpoint inhibitors is not the same as autoimmune hepatitis. Autoimmune hepatitis patients tend to be younger. They tend to have other autoimmune diseases. Their bilirubin and IgG level is higher. They're more likely to require more immunosuppression and a longer time to see improvement. In patients with immune checkpoint inhibitors, we can see that they can have a low ANA, but they tend not to have smooth muscle. And on their biopsy, they do not have evidence of plasma cells. So we really need biomarkers to really help us know if we can predict and detect early these toxicities. Some of the things that have been shown is if you have a presence of an autoantibody, this can predict the development of an adverse event. If there is a baseline low neutrophil lymphocyte ratio, this correlated with higher rates of immune-related adverse events. And interleukin-17 was correlated with risk of diarrhea and colitis. There has been a profile of 11 different cytokines that has put together called cytotox that can be used or is being looked at in terms of a predictive tool to look for chances of developing adverse events. I also wanted to touch on just, in terms of PD-1 and safety and toxicity prior to liver transplant, patient are using this to downstage transplant. We're seeing some remarkable results. This was from a single center at Mount Sinai. There were nine patients. They were listed. They underwent transplant after being treated with a PD-1 inhibitor. The majority occurred within four weeks of their last dose. They did have overall very good outcomes. However, there have been single reports of patients having disastrous loss of graft and death related to the strategy. So we certainly need to know the optimal timing of when patients could potentially receive this prior to undergoing transplant. This I just wanted to highlight was from Nature Reviews. It is a very good guide and it highlights all the different things that I've talked about in terms of diagnosis, the epidemiology, the management, et cetera. So I just wanted to point this out for your use. And I also wanted to, lastly, we didn't talk too much about Avastin, but the risk of bleeding. And in the IM-BRAVE trial, patients had to undergo an endoscopy prior to enrollment. There were 26% of patients who had baseline, paracese, 11% in the atezobev were treated versus 14% in the sarafnib. There was a slightly higher risk of upper jaw bleeding in the atezobev, 7% compared to 4.5 in sarafnib. So this I would consider toxicity and needs to be managed, but can be done so easily with endoscopy. So my takeaway points are there's a higher rate of hepatotoxicity in HCC than non-HCC. The treatment of immune checkpoint inhibitors is complex and complicated and requires multidisciplinary team. Recognition is important in order to start the appropriate treatment and biomarkers are needed to predict and guide therapy. Thank you for your attention.

hepatocellular carcinoma

combination therapies

molecular classifications

genetic alterations

sorafenib

atezolizumab

Bevacizumab

clinical trials

sequencing active drugs

immunotherapies

checkpoint inhibitors

VEGF inhibitors