Good morning, everyone. I'm David Kaplan from the University of Pennsylvania in Philadelphia. And it is my great honor to introduce today's speaker, Dr. Josep Maria Llovet, a transcontinental translational researcher and investigator who's played a major role in transforming the landscape of hepatocellular carcinoma care worldwide. He is currently director of the liver cancer program at the Mount Sinai School of Medicine and director of translational medicine at the University of Barcelona. Like Leon Schiff, who co-founded the ASLD, Dr. Llovet was a co-founder of the International Liver Cancer Association, of which he was also the president, and has played a major role in guidelines committees for the easel in liver cancer. He's published over 280 papers in the field of HCC, including pivotal trials that have shaped the current landscape of systemic therapy in this field, most famously the groundbreaking SHARP trial. He continues to actively translate seminal findings from his laboratory, identifying driver gene pathways in both hepatocellular carcinoma and intravatic cholangic carcinoma into the clinical setting. And for these efforts, he's received numerous awards, including the AACR Landon Award for International Collaboration and the International Hans Popper Award in 2012. So please join me in welcoming Professor Llovet to the podium for the 2019 Leon Schiff state of the art presentation entitled Molecular Pathogenesis and Management of Hepatocellular Carcinoma. Thank you. Mr. Chairman, organizers and colleagues, first of all, I would like to thank the president of ASLD for inviting me to give this 2019 Leon Schiff State of the Art Lecture on Molecular Pathogenesis and Management of a Pathocellar Cardioma. Reality, I mean, it's an honor for me to be here and to honor Leon Schiff, that was the first president of ASLD. Here you have my disclosures. And this is the outline of my presentation. I will talk about epidemiology, molecular drivers and targets of therapy, staging systems, management of early, intermediate, advanced HCC, and finally, I will talk about precision oncology. Liver cancer is the sixth most common cancer globally, the fourth leading cause of cancer-related mortality. Over 800,000 new cases of liver cancer were reported last year, and around 1 million new cases have been estimated by 2025. As you can see in this figure, the incidence, the prevalence, and the mortality of liver cancer mostly occurs in Eastern Asia and Sub-Saharan Africa. In the United States, we have nowadays 37,000 cases per year. Liver cancer is the leading cause of death in cirrhotic patients, and the incidence is increasing globally. This slide summarizes the incidence rates of HCC according to geographical area. I just want to point out that the incidence in North America now equals the incidence in Southern Europe, both for males and for females, with 10 cases per 100,000 in males, 3 cases per 100,000 in females. And this slide shows you the mortality trends of patients with different malignancies in the U.S. between 1990 and 2009, and as you can see, for the majority of solid tumors, there has been a significant decrease in the mortality, with two exceptions. On one side, we have melanoma, and on the other side, certainly, we have liver and bile duct cancer, with an increase in the mortality of 50%. This figure shows you the incidence and risk factors of HCC globally. Hepatitis B virus is the number one cause of HCC, accounting for 54% of the cases. Hepatitis C virus accounts for 31% of the cases globally, and is still the number one cause of HCC in North America and Europe. Alcohol is the third cause of liver cancer globally, and as you know, NASH is an emerging cause of hepatocellular carcinoma associated with obesity and diabetes, and as you know, 35% of the adult population in the U.S. is obese. There are also co-factors, I just want to point out, aflatoxin B1, that is a toxic that produces point mutations in p53, and is a co-factor, particularly in sub-Saharan Africa. Let's move to talk about molecular drivers. In the pathogenesis of liver cancer, 70 to 80% of the patients have an underlying cirrhosis, and therefore, this is a step-by-step process from cirrhotic nodules to low-grade dysplastic nodules that have, the cells have a proliferative advantage as a result of this regulation of some pathways here, and then there is a proliferative advantage in using high-grade dysplastic nodules and open hepatocellular carcinoma. Only one gatekeeper has been identified so far, third promoter mutations, in 6% of low-grade cirrhotic dysplastic nodules, and 20% of high-grade dysplastic nodules, and around 60% of open hepatocellular carcinoma. Conversely, there is 20 to 30% of the patients with HCC without underlying cirrhosis. Particularly, these are patients with hepatitis B virus infection, also NICE-related HCC, and less so hepatitis C virus infection. And there is a small proportion of patients, around 1-2%, with other etiologies, for instance, related with the infection of adenovirus 2, and also hepatocellular carcinoma progressing from adenomas, particularly with the beta-catenin mutation. This slide shows you the landscape of mutations for all tumors. Here, you have the hyper-mutated tumors, more than 100 mutations per tumor. Here, you have the tumors with a low number of mutations per tumor, pediatric tumors and liquid tumors, and hepatocellular carcinoma stands here with around 60 non-synonymous mutations per tumor. In this study, we collaborate with the group of Jessica Thuckmann-Rossi to publish the first genome sequencing in HCC in the West, 250 patients, and as you can see, the most prevalent mutations were TERT, beta-catenin, TP53, ARID1, and so on and so forth. And then, there is a long tail of other mutations at the level of less than 10%. As you can see, the most relevant mutations, unfortunately, are undruggable. And in this meta-analysis of close to 1,000 hepatocellular carcinomas, for which we had whole exome sequencing again, the most prevalent mutations, TERT, P53, beta-catenin, ARID1, are undruggable, as opposed to those with low prevalence, for instance, platelet-derived amplifications of VHGF-A or FGF-19 that are druggable. Overall, it is considered that 25% of HCCs at least have one actionable target. And how do we stand in the context of oncology? Well, here you have the percentage of druggable targets in different malignancies, and certainly, HCC is one of the tumors with less druggable targets, unfortunately. Aside of the mutation and activation of the pathway as a result of mutations, we also have activation of pathways as a result of epigenetic alterations. For instance, notch activation, TCF-beta, 17% of the cases met activation in 50% of the cases in advanced HCC as a result of overexpression of a parasite growth factor, and insulin growth factor signaling also as a result of overexpression of insulin growth factor 2. Overall, we have established molecular classification of HCC. We can divide the patients in the proliferation subclass and non-proliferation subclass. Within the proliferation subclass, we have a subgroup with progenitor cell markers associated with high levels of AFP and poor outcome. We have another group characterized by TGF-beta signaling. And then, in the non-proliferation class, 25% of hepatocellular carcinoma have this beta-catenin mutation that is very dominant and is, unfortunately, undruggable. This subclass is mostly associated with alcohol-related HCC and hepatitis C virus-related HCC, and certainly, it shows better outcome. We only have one drug associated with these subclasses, Ramucinumab. I will talk about later on. It's a monoclonal antibody against BHCF receptor 2 that have shown only to be effective in patients with high AFP that will fit in this progenitor cell subclass. Also, we have defined the immune class of HCC. This is a study we published in 2017. 24% of hepatocellular carcinomas belong to what we call the immune class. We consider them as hot tumor, characterized by enrichment of T cells and also by enrichment of signatures of response to immunotherapy, as opposed to 25% of the cases that belong to the immune excluded class, with a paucity of immune cells and enrichment of mutations in beta-catenin. This is a hypothesis. We think that these so-called hot tumors may be the ones responding to checkpoint inhibitors. Let's talk a bit about the staging systems in hepatocellular carcinoma. It is now 20 years that we published the seminal paper describing the BCLC staging classification. And this staging classification has been adopted by the SLD guidelines, by the ESL guidelines, and by some guidelines in Asia, including the Korean guidelines. It establishes different prognostic stages and allocates the treatment of the patients. If I have to choose one slide to summarize the advancements in the treatment of HCC, the met needs, I will choose this slide. You have the five stages according to BCLC. In blue, you have the natural history of the disease, survival at early stages without treatment, 36 months, intermediate stages without treatment, 16 months, and advanced stages, 6 months. With resection, transplant, and local ablation, we have been able to move that to above 60 months median survival. With intermediate HCC and chemoembolization, we are now experiencing median survivals of 30 months. And with all the molecular target therapies, we are in the range between 16 and 8 months median survival, depending on the cases. And in red, you have the met needs. We don't have, nowadays, adjuvant therapy after resection and local ablation. Still, after 15 years, we don't have any combination that is able to improve the outcomes of chemoembolization, and other challenges remain in the management of advanced HCC. Here, just to summarize and link you, the levels of evidence, high level, randomized controlled trials, meta-analysis, and the strength of recommendations, green, highly recommended, as opposed to orange, red, non-recommended. Moderate stays for phase 2 data. Low stays for retrospective studies. As you can see, sorafenib, lembatinib, rego, cabo, ramu are the five drugs highly recommended. Chemoembolization, radiofrequency, as well in early HCC, and liver transplant and resection. So, let's talk a bit about the advances in early HCC and intermediate HCC. Well, what have we learned in surgical resection? First of all, single tumors in patients with otherwise well-preserved liver function are the ideal candidates for resection. And here, in this flowchart, you can see that patients without portal hypertension and single tumors, as it is recommended by ASLD and ESL guidelines, have 70% five-year survival and very low risk of decompensation and very predictive mortality of less than 0.5. As opposed to what the Asian guidelines are proposing, the Asian guidelines are proposing to push the envelope to resect tumors in patients with portal hypertension, minor resections, one or two segments, or without portal hypertension, three segments. So, multinodular tumors here. And then, the risk of decompensation goes up to 30% and predictive mortality up to 9%. Another thing that we have learned is that anatomic resection is significantly better than non-anatomical wage resection, certainly because with the wage resection, you are leaving here satellites and eventually tumor in the small branches. A summary of this, how this translates in real life, in this NISA study, including 2,000 resections, those patients that are ideal candidates for resection have a five-year survival around 70%, significantly better than those patients that are ideal candidates, but actually were not resected, or those patients that were resected without being ideal candidates. Unfortunately, we don't have adjuvant therapies after resection local ablation. This is the STORM trial, the largest trial ever reported, including more than 1,000 patients, comparing xoraphanib versus placebo. As you can see, recurrence-free survival, 33 months for both arms. Therefore, in terms of adjuvant treatment, there are no recommended treatments in guidelines. Phase 3 trials are currently testing single agents, nivolumab versus placebo, pembrolizumab versus placebo, checkpoint inhibitors, or even actually combinations, atezopembro, durbolumab, bebecizumab. In terms of liver transplantation, here you have the hallmark study of Vincenzo Mazzafferro, published in 1996, establishing the guidelines for a transplant based on the Milan criteria. The outcome suspected at median survival up to 10 years and recurrent rate around 15% have been adopted by SLV, ESLV, CLC, and the Consensus Conference, and in this large meta-analysis of retrospective studies, including of thousands of patients, you can see that those patients, Milan in, are having a better outcome, hazard ratio 1.68, comparing to those patients with beyond Milan. Certainly, there have been several efforts to expand the criteria of transplantation. Here, I'm showing that the list of efforts, including the UCSF, the Up to 7, the Toronto proposal, and the Kyoto proposal. With these efforts, the outcomes reported are around median survival of 5-7 years, and recurrent rates range between 20 and 40%. And I have to notice that these recurrence mostly occur in extrahepatic sites in 60%, and both liver and extrahepatic sites in 40%. In this meeting, we have presented a multicenter study, including 2,500 patients transplanted for HCC in all these centers, 330 patients downstage to Milan, 113 beyond Milan, and more than 2,000 within Milan. And interestingly enough, with a very, very large follow-up, we can say that 10-year survival of patients within Milan is nowadays 61, and this may be the benchmark at this point, 61%, and 52% for patients downstage to Milan, significantly lower than the Milan, but still outstanding survival, and 39% 10-year survival for patients beyond Milan. And in the same direction, recurrence, 14% for patients transplanted within Milan, recurrence at 10 years, 20% in patients downstage to Milan, and 47%, which is unacceptable for patients beyond Milan. What about local ablation? Well, years ago, it was established that radiofrequency is superior to PI. Here you have five trials and the meta-analysis. And after that, other techniques like microwave, cryoablation, and others have been proposed, but none of these techniques have shown to be superior to radiofrequency. Therefore, radiofrequency is currently recommended as the front-line treatment in patients with less than 2 centimeters BCLC0, and also in patients BCLCA not suitable for resection or transplantation. Currently, there is an ongoing trial, the Optima trial, comparing ThermoDOF plus radiofrequency versus radiofrequency in phase 3. Chemoembolization, here you have the meta-analysis that in 2003 established chemoembolization compared to suboptimal therapies as the standard of care for intermediate HCC. During the last 15 years, there have been a lot of advancements in the technique, more selective treatments, and in the delivery of the drugs, the drug-eluting beads, beads loaded with doxorubicin or other chemotherapeutic agents. But no major advances have occurred in terms of the outcomes. Here you have the classical contraindication. I have to say that the applicability of TACE in BCLCB is only 50 to 60 percent because the other 40 percent of the patients either present liver dysfunction or technical contraindications. New efforts have tried to improve the outcome of TACE combining with either sorafenib, brevanib. Trials were negative, median survival in both cases beyond 26 months. So nowadays, still chemoembolization after 15 years is a standard of care for intermediate and other alternatives are currently explored. For instance, combining TACE with bevacizumab and durbolumab. What about Y90? Very promising. The first studies with branch portal vein mutation, sorry, branch portal vein thrombosis reported 16 months and 17 months median survival in patients with child book A class. And these were very appealing studies that induced the initiation of these six randomized control trials, mostly comparing Y90 versus sorafenib. What has happened is as follows. Three trials have been negative, three trials negative, no showing any superiority for Y90, and three trials have been halted. So unfortunately, based on evidence, there is no room at this point for Y90 for the management of advanced HCC. Let's talk about which is the current management of this disease. We have the seminal trial we published in New England showing survival advantages comparing sorafenib versus placebo 10.7 versus 7.9 months has a ratio of 0.69. This established sorafenib as a standard of care in frontline. In therefore a meta-analysis that we conducted including the SHARP and the Asia-Pacific trial, we identified that patients with liver-only disease and patients with hepatitis C virus infection are those that benefit more from sorafenib. Conversely, child book A class patients, of course, median survival 13.6 months, but Chalbee and Chelsea have very poor outcome with sorafenib, and it's not recommended. After that, I'm listing here the trials that have been reported. As you can see, the majority in red are negative trials. Only one trial in orange, lembatinib, for non-inferior to sorafenib, and three trials positive in second-line therapy. Lembatinib is a multikinase inhibitor blocking FGF receptor 1 to 4, show median survival 13.6 months and 12.3 months, and in the subgroup analysis mostly show that benefits patients particularly with hepatitis B virus infection and aggressive tumors. In second-line therapies, we have regorafenib that have a similar structure to sorafenib except for this fluorine, and the trial was randomizing regorafenib versus placebo in patients progressing to sorafenib. Median survival 10.6 months for regorafenib, 7.8 months for placebo with a hazard ratio 0.63, establishing regorafenib at that time point as a standard of care in second-line. The other drug that also showed positive results, a multikinase inhibitor blocking several kinases, including meth, that is disrupted at least in 50% of the patients advanced HCC, cabozantinib, 10.2 months median survival as opposed to an 8-month placebo that is what is expected in patients in second-line, and particularly cabozantinib again was favoring patients with extravagant spread, hepatitis B virus infection. Another drug is ramucirumab. This is truly precision oncology, is a monoclonal antibody against BHGF receptor 2, and just by blocking one target, the drug was shown to be effective in terms of survival advantages. This indicates the potency of BHGF signaling in the pathogenesis of HCC. Here you have the subgroup analysis of the first negative trial, the second trial just enriching the patients with FP more than 400, and the meta-analysis. Median survival here is 8 months and 5 months for placebo because we are talking about patients in second-line with very aggressive tumors, FP more than 400. So this is the landscape in front-line, lembatinib, significant for non-inferiority, and three drugs significant for superiority versus placebo, and this is the current landscape of management, sorafenib, lembatinib in front-line, rego, cabo, and ramu in second-line. What happens though with nivolumab and immunotherapies, checkpoint inhibitors? Well this has been a revolution in HCC, but also a particular revolution in oncology. Here is a very nice paper reviewing the most important treatments and the indications, and these treatments achieve between 15 and 50% objective responses. In HCC, this was the first trial testing nivolumab in first and second-line, 262 patients, objective response 16%, median survival 60 months, and those patients responding actually have an outstanding outcome with survival beyond 30 months. Unfortunately, PD-1, as you can see here, there were no differences in objective response, and this is an unmet need. This drug was moved to front-line, head-to-head to sorafenib in a phase three trial, again objective response 15%, and these are long-lasting response that happened with checkpoint inhibitors, but unfortunately the trial did not show survival differences as of ratio 0.85. Despite that, nivolumab in this trial showed the best outcome ever reported in front-line advanced, 16.4 months. The same happens with pembrolizumab in second-line, objective response of 18%, again long-lasting responses impacting in this population the survival, but considering the whole population has a ratio of 0.78, and the p-value, despite that, was 0.02, did not hit the pre-specified p-value of 0.017. So again, pembrolizumab here is showing the best outcome ever reported in second-line, and again, there is an unmet need to identify the responders. Here you have in front-line where nivolumab stands, so since the trial was designed for superiority, did not hit the endpoint because it's crossing the one, the upper boundary, and here with pembrolizumab in second-line, the upper boundary is below one, but they expand the alpha in other co-primary endpoints and in interim analysis, unfortunately. So can we identify biomarkers in oncology? Only two biomarkers have been validated in phase three studies, tumor mutational burden in non-small cell lung cancer and melanoma, and PD-L1 expression in melanoma. We describe, as I mentioned, hot tumors and cold tumors, potentially responding and potentially primary resistant to checkpoint inhibitors, and we recently reported this year that beta-catenin activation in experimental models promote immune escape and resistance to anti-PD-L1, and in a small study of 27 patients, certainly those patients progressing to checkpoint inhibitors had beta-catenin and Axin-1 mutations, and this needs to be validated. Finally, we'll talk very briefly about the combinations, combining drugs that are blocking TKIs, monoclonal antibodies, or receptors in the cell and then are blocking the immune system or enhancing the immune system. I'm going to talk about bevacizumab, atezolizumab, and also I'm going to talk lembapembrone. Well, atezolizumab, bevacizumab, with 73 patients, objective response by modified receipts, 34%, so the bar now is getting very high, got FDA breakthrough designation, and the trial was designed, and this has been a trial in which the patients have been recruited in one year, phase three, 500 patients, and a couple of weeks ago, we know that this trial is positive. Atezolizumab, plus bevacizumab, is superior to sorafenib, and this is the press release. We will know the data by the end of this month. The other combination is lembatinib, pembrolizumab, 67 patients, 42% objective response by modified receipts, also got FDA breakthrough designation, median survival 20 months in frontline, and progression free 9.7, and here you have the trial that is ongoing. And finally, two minutes to talk about precision oncology. So this is a very nice study with 10,000 patients running memorials on Kettering Cancer Center, all these patients were checked for mutations, 500 mutations, and actually, 37% of the tumors were able to be treated as a result of this analysis because at least they have one actionable driver, 37%. And this is the case, for instance, of erythema amplifications in breast cancer, treatment with rastuzumab, BRAF mutation, 60% in melanoma, treatment with bemoraphenib and ALK fusions, and chrysotinib in non-small cell lung cancer. All these mutations have been identified with biopsies of the tumors. So what do we have in HCC, I'm just showing this example. Over-expression of FGF19, 7%, over-expression, 20%. We have this drug, it's very selective, blocking FGF receptor 4 in PDX models working very well, decreasing the volume of the tumor, and as a result of that, a proof-of-concept trial was designed in patients with FGF19 positive and negative. And as you can see here, objective response was 16% in patients with FGF19 positive with this drug called Fisogatinib, as opposed to 0% in FGF19 negative. And this is a proof-of-concept we published this year in Cancer Discovery. So Mr. Chairman, in conclusion, the incidence of HCC is growing globally. We have identified molecular classes, the proliferation and non-proliferation class, enabling the understanding of the disease. We know the most prevalent oncogenes and tumor suppressor genes, unfortunately, these are non-targetable, and only 25% of the tumor harbor an actionable target. So there is an unmet need of translating knowledge into trials and management. I just put the example of Fisogatinib in FGF19 positive patients. In treatment systems, the BCLC staging classification is widely used in the West, because it links prognostic stages with treatment allocation. In early resection transformation, STILA recommended, intermediate chemicalization recommended, frontline advanced, SORAFINIB, LIMBATINIB, second line, REGO, CABON, Ramoserum mapping patients with FB more than 400. Checkpoint inhibitors provide clinical benefit, but probably only to a subgroup of patients. And we need biomarkers to identify this subgroup of patients. And combination therapies, for instance, a test of ebolism have recently been reported to be positive, and this will be changing the landscape of management for HCC. Finally, I would like to thank the members of the Mount Sinai liver cancer program, and my mentor Scott Friedman, and also other friends, Augusto Villanueva, that are sitting here in the audience, and Myron Schwartz, and also the members of the Translational Research Group in Hepatic Oncology in Barcelona, and all the members of the liver unit, EDVAPS in the Spital Clinic in Barcelona. Thank you very much for your help. Thank you.

Dr. Josep Maria Llovet

Liver cancer researcher

Hepatocellular carcinoma

Molecular pathogenesis

Systemic therapies

Precision oncology

Immunotherapies