Dear friends, I think we shall start. Hello and welcome to the final session on the second day of ITCS 2022. Thank you for staying here in the room. Also, my colleagues already used one half an hour of our time and you're still here. Thank you for that. But you are good people and I have a personal experience today. I'm proving that IGCS people are good people. Today I lost my two credit cards, the Visa and the MasterCard. And I was absolutely in horror. I went to the security and they asked for ID and then they gave me the two cards that somebody found it and gave it to security. And that's really a proof that you are good people and I'm very grateful that you did that. And my name is Andreas Dubois and I'm co-chair of this session alongside with Thomas Herzog, Amit Oza, Isabel Rey-Cocard and Anil Sood. I think it's best that you join the podium already to save some time. We will have two debates and each topic will have a pro and a con and let's see if we have time for a little bit of discussion. And then we have panel presentations and after that we will have Q&A and you may join the microphones that we can hear you. So the topic of the first debate will be Shapening the Scapelle with Molecular Biology. And Dr. Anil Sood from the Andy Anderson will take the pro and Dr. Florian Heitz from my institution in Essen will take the con. So Dr. Sood, please begin. Good afternoon everyone and thanks so much, Andreas. Really great to see everyone here in person for this meeting. So these are my disclosures. So I won't spend a lot of time on this. I think to this audience, you all know what our patients go through in terms of the different steps. One thing that continues to evolve is the upfront management in terms of who should have surgery. Should all patients have surgery or do you start with neoadjuvant chemotherapy? So there's still a number of questions that need to be addressed and obviously surgery can also potentially play a role down the road in patients with relapsed as well. But for today's purposes, I'll focus mostly on the upfront setting. And the main questions, again, that we continue to face are those that I list here. And in a lot of ways, we also need to understand what the biological differences are related to the patterns of disease. And I'll show you some examples of how we're trying to address these. So let me start this. And over the years, we've learned a lot about is ovarian cancer really all the same? There are those who historically, we used to lump them all together. And most of us did surgery to start out with in terms of providing care for these women. So that's a lumber kind of a view. But over the years, we've evolved into the splitter view, which is that, indeed, ovarian cancer is not one disease. And even within specific histological subtypes, that perhaps we need more personalized surgical approaches in terms of who might benefit the most in an upfront setting versus considering surgery at a later time. Now, this is a summary of some of the studies that have been done over the years. And over the last 50 years, we've evolved from considering, quote, optimal resection, meaning leaving less than a 2-centimeter residual at some points to less than a centimeter, to really thinking more about it in terms of complete gross resection. Some people call it R0, but really, complete gross resection of disease. And that proportion, yes, it can vary depending on the institutions and so on. But it's certainly not a very high number at most of the institutions. And the question comes up is that if you know that you can't achieve complete gross resection, are you better off starting with an alternate path, meaning giving chemotherapy, perhaps? There are many groups who've tried to predict who you can resect all disease in. People have used circulating markers, imaging, a variety of clinical parameters. There are lots of algorithms that have been developed. The point is that most of these are really not ready for prime time. And these, as a result, have really not been implemented. Today, I'll share with you, before I get to the molecular biology part, we needed a consistent way in which we take care of our patients. So 10 years ago, at MD Anderson, as a part of our Moonshot program, we started to use laparoscopic triage for those patients who were surgical candidates to try to make upfront decision making. And I'll show you some data just before I get to the molecular biology part. And we adopted the scoring system that Anna Fagoti had devised. We started for the first three years as a quality improvement project. So this was a QIB approved project where there were two surgeons who did the scoring because we really wanted to get some sense of how consistent the scoring was. After the three years, we realized that among our 24 G1 oncologists, there was about 97% concordance in the scores that would affect decision making. So we went to single surgeon scoring after the first three years. Now, the triage worked as follows for the first three years that those patients who were surgical candidates, they had laparoscopic assessment. And those who had a score of less than eight from that scoring I showed you on the last slide, they underwent upfront debulking surgery. And our complete gross resection rate has been close to 90% throughout this period of time over the last 10 years now. And those who had a score greater than or equal to eight, they had neoadjuvant chemotherapy. And as long as there was some degree of response, then they would undergo interval tumor reductive surgery. And our CGR rate is about 76% in that setting. This is the PFS data that Dr. Nicole Fleming presented a few years ago at SGO. And this has also been published. Clearly, these are very different patients in terms of their outcomes based on R0. So the dark blue line is R0 upfront. The dark red line is R0 or CGR at interval surgery. And the worst outcomes are for those patients where you leave residual even at the interval tumor reductive surgery. So with this as background, we also wondered that whether this is a cost effective way of going about this. So there have been at least two studies that have now looked at cost effectiveness of this kind of an approach. And both really showed that a laparoscopic assessment reduces the number of futile laparotomies. And indeed, it was felt to be cost effective. Now, let me shift gears and discuss with you how we've tried to look at the molecular biology and what role this could play. Again, the same algorithm I've outlined on the left-hand side. We really wanted to do a very in-depth understanding of the biology. So three groups of patients, those who had upfront debulking surgery and had CGR, those who got neoadjuvant chemotherapy but had excellent response to chemotherapy defined as virtually complete resolution of disease, normalization of CA125, et cetera, versus those who had relatively poor response to chemotherapy, meaning they had essentially stable or minimal response to their treatment. These are all baseline samples. We looked at primary site as well as two to three metastatic sites. All of these samples were subjected to RNA sequencing, whole genome DNA sequencing, as well as the NGS T200 panel, full proteomics that was spatially defined, as well as immune profiling. And then we did extensive integrative analyses. Today, I obviously won't have time to go through all of the data, but here's a snapshot. The oncoplot on the left-hand side are the primary sites and to just to the right-hand side are the metastatic sites. Again, there's a lot of data and differences in these populations, but summary is that even, for example, P53, which is frequently mutated in high-grade serous cancers, nonsense mutations in P53 really were detected exclusively in the neoadjuvant group. Moreover, CSMD3 and PIK3CA, the nonsense mutations, again, occurred exclusively in the poorly, poor responders who got neoadjuvant chemotherapy. Another example is these are copy number variations. Again, same pattern in terms of how the data is reported. CGR on the left side, neoadjuvant, excellent response in the middle, and poor responders on the right-hand side. The one that really stands out here is NF1 loss, really remarkable loss rates in those who had CGR up front compared to the neoadjuvant populations. We're pursuing this further for functional validations and drug development purposes. You know, with advancing technology, we've also deployed spatially resolved transcriptomics, meaning that each cell or each two to three cell area we have done full RNA sequencing on, and this is done on the slide so you don't destroy the spatial architecture. The bottom two slides just simply show you the map that reflects the spatial analysis. This we just reported earlier this year. Now, those who have poor response to chemotherapy, they tend to be stroma-dominated cell groups. A lot of EMT signatures vary consistently in this spatial analysis and higher expression of DNA-repaired genes. Those who have excellent response, they tend to have much more prevalent immune-related areas. I'm going to shift gears and show you another way to look at this. You know, most gYN oncologists realize that, you know, you can look at these two slides here. They're both high-grade serous, but when we look at the gross disease in the operating room, many times it looks very different, and yet we haven't really formally recognized or analyzed it in any way. So all the patients in the moonshot who undergo laparoscopic assessment, we have video-recorded data on all of these patients. So we analyze this, and by the way, these are then the morphology patterns that you see intraoperatively, even though the HND sections look virtually identical. So with many of our colleagues, we developed what we are terming for now type 1 and type 2. For example, type 1 is deep infiltrative appearance. Again, most gYN oncologists will recognize this kind of a pattern. Type 2 is a superficial appearance with exophytic nodules and so on. We list this all here. This paper will be out in the next two weeks. Now again, we've done a fairly in-depth analysis of these two morphological patterns of disease. Type 1 tends to respond better to neoadjuvant therapy. However, at a molecular level, you see much higher EMT signatures, hypoxia, angiogenesis, and so on. Type 2, the more superficial exophytic disease, tends to have clinically, I'll show you some of the clinical data, but it tends to have much more of mixed signaling, cell cycle progression, and so on. So here's some of the clinical data between these two patterns. The type 2 disease tends to have longer operating times. They tend to be somewhat more bloody cases in terms of EBLs. The complete resection rates and so on are relatively similar, even though there are differences in their outcomes. When we did full proteomics analysis on this, this is a PCA analysis to just show you that indeed these two patterns of disease diverge, and some of the pathways that I alluded to in the earlier slide are reflected on this heat map, as well as the PCA analysis. Moreover, we did a spatially oriented metabolomic analysis using daisy mass spec, really fascinating data where we can look at metabolites both in the epithelial and the tumor compartment, as well as in the stromal compartment. The brief version is that type 1 disease tends to have much more in terms of arachidonic acid, fatty acid type of metabolites, whereas type 2 tends to have cardiolipins, as well as phosphatidyl inositols. There are a lot of other differences that, again, because of time, I won't be able to go into the details. So, in summary, the laparoscopic assessment that I shared with you is indeed effective and enables personalized surgical approaches, and over the last 10 years, it's been a very consistent way that we approach. We've had a 98 percent compliance rate as a part of this QI project at MD Anderson, so we really follow very similar patterns in terms of practice. There are extensive molecular and cellular variability in high-grade serous disease in clinically defined subgroups, which is what I shared with you, and there's enormous heterogeneity even in morphology at a gross level in this disease. So, in a lot of ways, you know, the question isn't as much as how do we sharpen our scalpel, but it's also about, as GYN oncologists, you know, are we paying attention to what these subgroups look like, what the in-depth analysis look like, and are we open to adapting our own practice patterns, and how do we use this information in a smart way to personalize care for our patients? Thank you. Thank you. Thank you, Dr. Sud. I thank my colleagues and collaborators. Now, Dr. Floria Nights will take the con, and hopefully we will have some questions after that. So, thank you very much for the invitation and this great talk. So, basically, I would like to acknowledge that I'm, of course, also interested in the progress, medical progress, and also in analysis, taking molecular biology into account. Nevertheless, I believe I found a couple of issues which might be difficult to deal with in this respect. So, I found some things where we can discuss whether this is really a good idea to use molecular biology for sharpening the scalpel. My conflict of interest, which are all out of the range of this talk. So, if we operate on patients with advanced ovarian cancer, we offer the patients the option of improved survival. So, if we reach tumor residuals from 1 to 10 millimeters, we have a short, a small, but significant improvement in patients with high-grade serous, but also in patients with low-grade serous ovarian cancer. And if we get to complete resection, we really get a high frequency of prolonged survival. And even if you see here all these lines, sometimes we cure the patients. So, we operate on the patients because we have a cure in a considerably high number of patients. If we go to expert centers treating patients with ovarian cancer, we can see that approximately two-thirds of the patients will be completely resected. And in 25% of patients, residual disease between 1 and 10 millimeters will be apparent, which we would say that there is some benefit of primary surgery. In only 9% of patients, residual disease larger than 10 millimeters will prevail. And this is, of course, a group of patients where we would think that they do not benefit from primary debulking surgery. In addition, there are some patients who have a 30-day mortality or will not start chemotherapy. So, in the end, we have 88% of patients where we would assume that they have a benefit from primary debulking surgery. So, basically, the probability to achieve a positive or a non-negative outcome in expert centers by PDS is very high. So, how could it look? So, Anil showed us one way to get the idea how a tumor is localized in the abdomen. Another way would be that you take a biopsy and then do the histopathological examination. And you would go for some kind of test for a gene signature, for example. And then you would say, okay, you do a primary debulking surgery, or you would do an alternative approach. The question would be, how long does a test need to be conducted? Some days, some weeks. If we come from the HRD test, it takes four weeks. So, you have a symptomatic patient, and now what do you want to do? You want to say to the patient, no, we wait until the result before doing anything or starting new adjuvant? I don't know. The other thing is the alternative approach. So, interval debulking surgery after new adjuvant is not so easy to discuss shortly. However, if you see patients with primary ovary, with PDS, and complete resection have the best survival. However, and I think this is really important to acknowledge, that patients who undergo interval debulking surgery, they have to be completely resected. Patients in whom tumor residuals will be available will be seen after new adjuvant. They do not benefit from any surgery. So, this is really a big issue. On the other hand, if we look at the data from GOG 152, if you are in an expert center and you can't operate on the patient primary, you can't operate on the patient later on. So, is it really a good idea to have an alternative approach at this time? I don't think so. Let us explore a little bit the domains why patients might be not operable. So, basically, we believe that there are three domains. There's inoperability, which means something like the patient is in a bad condition, medical problems, so we can't give really radical surgery. And then we have this flurry of things about non-resectability and insufficiency. If you look here on the right side, there are some papers which described why they had, at what locations in the abdomen they had residual disease. And of course, everybody knows the mesenteric diffuse bowel involvement with miliary disease. This is really a big issue. Nevertheless, if you look, for example, for diaphragmal carcinomatosis, this is described to be a factor for residual disease in 74 or 51% of patients. This is really a problem of surgical capacity, as well as, for example, bulky lymph nodes. So the different items of structural deficiencies, so non-resectability versus insufficiency, obtaining complete resection in a high proportion of patients makes it difficult to identify a molecular marker to distinguish between those both. So we would like to distinguish or to figure out only the patients where we would let tumor residuals up to one centimeter. If we go to the molecular analysis, which have been published in the last years, this is mainly based on meta-analysis with approximately 12 trials, which have been working with gene expression analysis. And then here, it's combining the TCJ with the Totfield data set, and they did some biostatistical analysis, and they found a debulking signature with a PPV of 86%. And in another trial here with this meta-analysis, TCGA data again, it's another publicity-available data set. They combined these four genes, and they found an AOC to predict surgical outcome with an AOC of 0.733. Finally, I think the largest meta-analysis, where 12 available gene expression profile studies were included, tried to evaluate also the impact of a molecular marker. And sometimes it's really interesting to go into the method section of a paper to understand what they did. And this is pretty interesting. They said, accurate prediction of debulking standard proved to be difficult with an overall AOC of 0.59. Within our debulking signature, we selected seven highly differentially expressed genes with a known biological role in ovarian tumor genesis. And finally, they used this seven genes, and they found a model with an AOC predicting residual disease with 0.76. So a meta-analysis including 12 independent data sets is overruled by a distinguished guess, and the PPV to predict residual disease by molecular biology is lower than the PPV to predict residual disease in an expert center. So we have complete resection rates up to 88%, and here we have something about 70 to 80%. If we go into the data of this meta-analysis, they were highly heterogeneous, and even though if we go for the basis of this meta-analysis of these 12 papers, there were very high heterogeneity with respect to residual disease with the location where the tumor was gathered, tumor material, gene expression analysis platforms, and so on. So here we have really a big, yeah, a big mess of data, and it's very, very difficult to find here a good validity of this data. So basically what we did then was that we wanted to look whether in an independent data set, if it's possible to find a new debunking signature, and we thought we take the data which have been published earlier, which I showed you now, and we wanted to try to validate anything of this in addition to the TCGA where we thought that this might be a good idea because there were some data out saying that myosincomal subtype patients would have a worse resection status. Finally, our biostatisticians said we do some machine learning, algorithms, and so on. If we do this, we find, if there is anything, we will find it. So and what we did was, what we found was that we had here AOCs of something about 0.5 to 0.62. So really, really big problem with validating. None of the data which have been published could be validated, nor could we find something new. So finally, no validation of previously published data in the independent data set. So I think this is a robust situation to say we do not really sharpen the scalpel with the data out there at the moment. Nevertheless, what is the idea? How could we do better? I think we should not use tumor biologic markers to hide surgical insufficiency. It's very, very important to learn the fellows to use the right side of the scalpel. We don't have, it's not reasonable to have a residual disease, for example, at the diaphragm. Nevertheless, identification of patients who don't benefit from PDS is a very important unmet medical need. But as I told you, this is not only the problem of resectability, but also of the patients who are not fit enough for debulking, which will have severe problems afterwards. And I think Anil showed it very nicely that if we want to go further with analyzing tumor with the idea to identify patients upfront in whom primary debulking surgery is maybe not the right way, we should use really very good samples from centers where the complete resection rate is high, which means that we have a low frequency of surgical or logistical incapacity, and so that we learn more about the pathogenisms of residual disease. So thank you very much for your attention, and I'm happy for the discussion. Great, great. Dr. Hertz, please join us here at the podium. We have time for one or two questions, although we are late in the evening. Yeah. Yeah. Good evening. Both the presentations were, of course, excellent. I have a question for Dr. Anil. So whenever we do these interval debulking surgeries, how do you tackle with extreme desmoplastic reaction? And doesn't that kind of leave the surgery incomplete nevertheless? And wouldn't such a patient be better off if he were to do a primary debulking surgery? And if I might add, is there any way to even predict that which patient is going to actually have so much of desmoplastic reaction after they have received chemotherapy? Yeah, all important questions. You know, in an interval setting, you obviously see a fair amount of variability in terms of, number one, what the patterns of disease look like, but also I would say that at times it can be very difficult to distinguish what's actually disease versus scar, or what may look like disease but ends up simply being either a fibrotic response and so on. So Dr. Fleming is actually leading an intraoperative effort where we use what's called a mass spec pen, which gives you real-time feedback, essentially in 15 minutes, what the metabolite profile looks like. In terms of resection, you know, the percentage of patients that we leave residual disease at interval overall is about 20 to 25 or 24 percentage. So it's not a real high percent, and these are obviously incredibly difficult people, patients up front, where it would be virtually impossible to achieve a complete gross resection. But I think in terms of predicting who does or doesn't get a really hardcore desmoplastic response, I don't know that we have great understanding of what that actually reflects. You know, we do have some knowledge of, you know, TGF beta, perhaps CTGF, that I think is gonna come up later in this session as well, that may play a role in that kind of biology. Antonio. So, Anil, I have a question. If I understood well, you saw that there is a PA3K mutation in patients with neoadjuvant chemotherapy and residual tumor after the surgery, but that was not seen in the primary debulking surgery with residual tumor. So a couple of things. First, what is your interpretation? What is the biology behind that, that the correlation between the residual tumor after surgery and PA3K mutation? And second, could be that an opportunity for a trial in that setting with a PA3K inhibitor or something like that? So, Antonio, I think that, again, great questions. You know, in some ways, it goes to, first of all, trying to understand, you know, these are clinical subgroups that we're starting off with in terms of what the differences are. I think to commit to a given drug or given pathway, we would obviously need to do biological validation of how robust those kinds of combinations may be, but that's really the intent of these patients who tend not to do well in terms of what else we could combine and PA3 kinase would be just one such pathway. I think there are certainly other druggable pathways that we should consider. So I think we, is there a question in the back? Oh. Yes, thank you. Thank you for a great presentation. I was wondering if you can comment on the turnaround time of getting the molecular signature and whether it can be utilized for clinical decision-making. I think, yes. The turnaround time, and Florian has addressed it. What is your turnaround time for your molecular profile? You know, I think it depends on what molecular profile one contemplates in this setting. You know, Dr. Shannon Weston actually has a new adjuvant trial, for example, with PARP inhibitor where the turnaround has to be rapid because it involves decision-making in terms of mutation assessment. I think that depending on what assay the data I shared with you really cut across the spectrum of many omics technologies. Ultimately, the question is, you know, what do you wanna use it for? As I think both Florian and I would agree that it's not really ready for surgical decision-making at this point because then we also have to consider, you know, what are we biopsying? And there's clear heterogeneity or differences whether you take a sample from the ovary or the omentum or, you know, whatever other site there may be. But if it's a decision-making related to the question that Antonio asked in terms of perhaps picking drugs or that are pathway-oriented, that could be done. And the technologies are moving at a pace where, you know, the turnaround time can be reasonable for clinical use. But I think just to add one thing is, I think a very critical point as we see from the problem right now with HRD testing is we need really a lot of tumor. So I believe that this 12-Gauche biopsy needle things don't really work good. So if we do something, we really need a good amount of tumor tissue to do this kind of analysis. And that speaks for the laparoscopy in that case or an exploratory mini-laparotomy. So I think we learned that they both agree that we currently don't have a lot to sharpen the scalpel so far, but many groups are on the way and we hopefully get some additional evidence for helping us to decide. I think we continue and we now come to potential impact on standard of care, review of pending trials, moderated by my friend, Emmett Osa. So please take over. Thank you so much, Dr. Dubois. We're going to sort of now go on to the next panel. Now, first of all, I'd like to welcome Dr. Isabel Rey-Cocard to actually start giving us a talk in terms of potential impact of standard of care and a review of pending trials. Thank you so much. I would like to thank the scientific committee to invite me to speak about this ongoing and future clinical trial in the first-line setting. This is my disclosure. As we have seen in the previous session, also carboplatin plus paclitaxel as a standard of care, we have only PARP inhibitor and BEZ in the first-line setting, and also PARP works very well in this population. When we look to the parallel one, we see that 50% of the patient, at the end, we will have. So the work is not done, and we need to continue to work. The question that we can address today is to work around the current standard of care, but also to move to innovative treatment as immunotherapy or novel agent. And I will speak a little bit about a particular histotype. If we remember the last conference of consensus from the GCIG, we have mentioning that for neoadjuvant chemotherapy, for example, we need to conduct the interval debulking surgery after a three or four cycle of chemotherapy. And there is also a question around the IPEC use. If we go back to the neoadjuvant question, and we discuss again about this point, what we can obtain if we increase the number of cycle of chemotherapy, specifically in the high-grade serous ovarian cancer, where we know that there is a lot of sensitivity to platinum-based chemotherapy. Unfortunately, a recent publication from France report that delaying the chemotherapy, adding more cycle, will not be so much able to increase the overall survival of the patient. And so it's time to ask the question. We are currently leading a randomized phase three trial asking this question in collaboration with the Brazilian group. And this remain an interesting question dedicated to high-grade serous ovarian cancer. The question of IPEC is not already answered. We have first randomized phase three trial, we put benefit, we need to have a second one to confirm the results. And this one is ongoing in the first line stating, including patient with stage three or selected stage four. And the patient will randomize at the time of the upfront surgery with no residual disease to receive or not IPEC therapy. The question about the current standard of care, carboplatin paclitaxel every three weeks was also explored several times looking to the weekly paclitaxel use compared to the three weekly. If we have seen some results in the Haasian population right now in the Caucasian population, we are not sure that we improve the PFS and the overall survival, looking to the results of the recent randomized trial. The question is also, what about when we had Bevazuzimab? And this question is ongoing with the ICON8 program, and we will receive the results at the end of the year or next year. Looking to the recent results with the PARP inhibitor in the first line setting and the confirmation of the overall survival benefit using PARP inhibitor, some of our colleagues and we have seen the debate before asked the question about the Bev looking to the good results of the PARP inhibitor. We have ongoing a phase three trial leading by the GYNECO in this particular population of patient with a stage three with no residual disease upfront surgery where we ask the question of the Bevazuzimab adding to Nihaparib. And there are also the same randomized phase three trial lead by the AGO group asking the question in the old population where all patient receive Nihaparib and the experimental arm will be Nihaparib plus Bev. So the point of discussion for the first line setting looking to our standard of care is really about the duration of treatment. If we have seen that Bev did not need to be prolonged, we can discuss about the PARP inhibitor. We have now two years versus three years. Perhaps we can move to increase the duration. We speak retrospectively about the PFS, overall survival and PFS2 as interesting endpoint looking that we have seen with the MITO16 where we know that the Bev does not seem to modify the efficacy of the following treatment. We need to do the same with the PARP inhibitor. There is some question about the crossover and I agree that perhaps it's time to include the second line therapy to explore the sequential treatment. But also we have to remember that safety and quality of life, also we are using maintenance therapy, need to be developed. Considering the immune therapy and novel agent, we have several ongoing large phase three who will be interesting to report some results. If we would like more, if we move in the detail of this current trial, we have the duo trial exploring the combination of Bevacuzumab, Durvalumab and Olaparib in the first line setting, including non-BRCA mutated population. This trial include 100% of patient with Bevacuzumab. It is a trial we as enriched in the HRP population. And finally, the question is PARP inhibitor plus checkpoint inhibitor plus Bev versus checkpoint inhibitor plus Bev versus Bev. This is an interesting question and we are waiting for the results. The second asking the same question, but in this particular trial, we are exploring the Pembrolizumab instead of the Durvalumab. And the question about the Bev is optional. There is no Bev moratory. But again, the question is PARP inhibitor plus checkpoint inhibitor versus checkpoint inhibitor versus surveillance. We are not well known about the use of Bev right now. It is also a population enriched in HRP patient. And the question will be about the power in the HRD population, the absence of PARP inhibitor in this population. And we have one arm where perhaps the patient did not receive maintenance therapy. The third one is an adaptive model using the Starlimab and niraparib, sorry, in the first line setting. In this trial, we have used the result from SOLO1 and PAROLAR1 to change the control arm. So this trial include patient receiving BEV or not. It is optional, and at the end, 50% of the patient receive BEV. All the patient in the BRCA mutated population receive a PARP inhibitor, and quite 75% in the wild-type population also. And this question is not exactly the same because looking to the number of patient who receive a PARP inhibitor, the question will be PARP inhibitor plus checkpoint inhibitor versus PARP inhibitor plus or less BEV. It is not exactly the same, and we can ask about the HRP representation in this trial. The last one is a real switch-maintenance question where the nivolumab was had to the first line setting just in the maintenance therapy and not in combination with the chemotherapy. There is no BEV in this trial. There is quite a well-balanced population in term of BRCA, non-BRCA, and HRP, and the question will be, again, PARP inhibitor plus checkpoint inhibitor versus PARP inhibitor. So we have some question about the different trial because finally the population is not exactly the same. We have some who increase a little bit more the population of HRP population compared to FIRST and ATHENA where probably we have more BRCA and HRD patient. And so how the group of HRP or HRD subgroup will be sufficiently poor to confirm the potential benefit when PARP inhibitor is not part of the control arm? How we will be able to conclude in the HRD-positive population? When PARP is part of the control arm, how checkpoint inhibitor will be able to fight the high benefit provided by PARP inhibitor in the first-line setting? And how is the choice of BEV or not will impact on the conclusion? Concerning novel agent and novel mechanism, I would speak to start with the oreogovumab. It is not a novel agent. We have several phase III. Unfortunately, we did not report benefit to use this antibody in the maintenance setting. But the good news in the randomized phase II, adding to chemotherapy, we see very interesting results. And so we are currently waiting for the confirmatory phase III trial, adding this compound to chemotherapy. We have to say that there is no BEV or PARP inhibitor in this large phase III. The other agent interesting to explore is the VGIL, that it is a compound using the autologous tumor cell. We are able to print the lymphocyte answer. And we have seen in the randomized phase II trial using this compound versus placebo after chemotherapy, again, no maintenance in this phase II trial, a huge benefit in the HRP population, where we have seen with several randomized phase III trial how this population have a worse prognostic and perhaps it could be interesting to confirm these results in a large phase III trial, including only HRP population. Some particular histotype I would like to speak about the low-grade serous carcinoma, and perhaps it's time to anticipate this escalation. It is what the NRG tried to explore in a large phase III trial, exploring a combination of chemotherapy followed by maintenance lethrozole versus lethrozole alone. This is an ongoing important phase III trial. To conclude on the first-line treatment, if we would like to have some discussion around the question by molecular subgroup, I would say that in the HRP population, we really need to know if the checkpoint inhibitor is able to change the story. And if it is not the case, we need to explore novel agent. In the HRD-positive BRCA wild type, for sure the question of the BEV plus PARP versus PARP is a question, and the question about checkpoint inhibitor able to increase the response to chemotherapy and to synergize remain an important question. In the BRCA-mutated population, it will be hard to beat PARP plus or less BEV, and probably the question of the duration of treatment will be at the next point. Considering other challenge in general in the first-line setting, we already speak about primary endpoint, PFS versus overall survival. Looking to the treatment in the second-line treatment, I think we need to continue to explore time to second subsequent treatment, but also PFS too. We are researcher. We need to know what's happened after the first line. And finally, because we introduced maintenance therapy in the first line, but also in the second-line setting, we need to continue to explore the quality of life of our patient, and perhaps to fix the duration of the maintenance therapy. Thank you very much for your attention. Thank you very much, Dr. Ricocart. That was an excellent presentation. Now I'd like to invite Dr. Brad Monk to do an approach in terms of looking at the analytics related to pending trials. Brad, take it away. Thank you. Thank you, Dr. Rosa. Thank you, Isabel, for a wonderful talk. What we decided to do was sort of divide it up between front-line and recurrent. There are so many trials, and I'm so excited to be here, and these are my disclosures. You know, we have PARP, obviously. Bev approved. It's the only targeted therapies in ovarian cancer. And obviously, chemotherapy, platinum, and taxane. And we use them over and over again. We do have liposomal doxorubicin and topotecan also approved. Gemcitabine, although not approved, is commonly used. So this idea of recurrent ovarian cancer being an unmet need is real. And just look, platinum-resistant recurrent ovarian cancer is undefeated. Undefeated for the last eight years. We have studied a number of agents which have not been successful, and I'm sorry, and we're all sorry, and most importantly, it's been very difficult for patients with the platinum-resistant recurrent cancer to survive for any period of time. And for us to palliate her symptoms. But we're learning the science. We're getting better, we're getting smarter. So I don't want this to be a discouraging message, but hopefully an uplifting message because I think all of you, all of us, are collaborating and working very hard. So first of all, as an international meeting, I want to call out the Gynecologic Cancer Intergroup, the GCIG. They meet twice a year, they last met in June. Chair is Dr. Okamoto and Ian McNish. This is the best list of ongoing trials in ovarian cancer. You can go on their website, you can download their slides. It's a wonderful resource. And I don't have time to go through all of those trials, but I want to make a few highlights because again, I think we're leveraging the science and I want to just touch briefly on four recurrent ovarian cancer strategies. Immunotherapy trials, again, with regulatory intent. There's lots of smaller sort of proof of principle studies, but I want to talk about studies trying to get it in the clinic. Late-stage clinical trials investigating antibody drug conjugates, probably the next class beyond chemotherapy, Bev and PARP and IO, now ADCs, and then some really novel mechanisms. And I want to thank Dr. Ray Cucard for helping me with the Bokeh trial, which I'm very excited about, so I have to end on that note. So I just wrote these. You mentioned the very nice frontline four trials that we're doing. There are also platinum-sensitive and platinum-resistant. I'm not going to talk about them. I was disappointed to hear that Adelante was negative, but we'll keep pushing and keep battling. So the first thing I want to talk about is novel immunotherapy. NIMVLUKIN. NIMVLUKIN is an engineered IL-2 cytokine that preferentially binds to the intermediate IL-2 receptor. It's the high-affinity receptor that causes most of the toxicity. So binding preferentially to the intermediate affinity receptor, we might get the therapeutic benefit and synergy that IL-2 and checkpoint inhibitor PD-1 affords. So this is being studied in multiple tumors. Particularly ovarian cancer, Artistry 7. And I'm so excited to see my NGOT friends and my Canadian friends and South American friends and Asian friends who are working together on these clinical trials. This is Artistry 7, taking physician's choice chemotherapy, gemcitabine, paclitaxel, liposomal doxorubicin, and topatecan versus the combination of NIMVLUKIN and pembrolizumab. So there are some small non-analytic single arm of NIMVLUKIN, Pembrolizumab to sort out the relative contribution. Thank you for doing this. Thank you, Dr. Herzog, for being a global PI of this, and I know we're very passionate about it. If you haven't heard about antibody drug conjugates, you haven't been at this meeting. Apifidimab, Rilozadotin, UPRI, is very novel and unique in its mechanism of action, in its target, NAPI2B. It has a very novel linker that sort of preferentially is inactivated and keeps it in the tumor. It has a novel cytotoxin, orastatin. The target, NAPI2B, is relatively high, okay? Two-thirds of the patients with high-grade serous ovarian cancer have it, and we use the TPS score greater than 70%, 75%. So you're familiar with this. This has been an important opportunity. What we do now is we have to optimize the dose. If you come to the agency, the FDA, and say, look, I got a medication that works with some toxicity, the first thing they'll say, well, how do you know that a lower dose isn't just as good? And you don't know, so you have to do the trial, and we sort of optimized the dose, started at 43 milligrams per square meter, but realized that maybe because it was working that we would keep the medication start at a lower dose, and you saw this with norepirib. We started at 300, and we optimized the dose in the PREMA trial. So you can see in sort of the brownish colors, the 36 dose, in the blue, the 43 dose, there's still responders, but obviously, less medication is less toxic, and you're gonna see this sort of evolution in our clinical trial program. The adverse events associated with this medication are very unique, because if you can, obviously, there's no ocular toxicity, okay, which we see with some of our antibody drug conjugates. There's no severe neutropenia or peripheral neuropathy, and even minimal alopecia. There was some pneumonitis early on, but as we reduced the dose from 43 to 36, there was less pneumonitis. So the dose moving forward, and thank you to Debbie Richardson for her work on this and others, the dose is 36. So you did a pivotal cohort with at least 100 patients, because that's what it takes to get accelerated approval that were biomarker positive, and it's basically finished enrollment. So that puts us in a good position here within the next, I'm not sure the timeframe, but maybe 18 months or so to get this in the clinic. Okay, so thank you for getting this done. We enrolled all comers. So the total enrollment has not been disclosed yet, but it's probably closer to 180 or so. So this is now done, and we'll have to wait to see what it shows. When you get accelerated approval in those single-arm cohorts, you have to do a confirmatory trial. We like to do them in an earlier line of therapy, and that study's called up next, and this is a maintenance trial of UPRI in that setting, and now we're starting to screen patients for this. Like our other studies, this is a GOG-NGOT collaboration, and thank you. We also combine novel combinations. So accelerated approval, earlier line of therapy, novel combinations, and then fourth in the life cycle management is other tumor types. Oh, that's happening too, but we're not gonna talk about it. So these are combinations called upgrade. The other antibody drug conjugate that's very proximal is mervituximab, sore of entancing. This now targets folate receptor alpha, very different cytotoxin, DM4. It's an antitubulin medication, maybe a more validated target, and it, like UPRI, has a bystander effect. Again, single-arm cohort. You already saw it, right, at the SGO, Soraya. About 100 patients, okay, in a high unmet medical need. These are patients that have failed bevacizumab, that if they're BRCA have failed a PARP inhibitor, and this is what we got. It's what you got. This has been filed to the FDA. This has an action date, the last week of November, maybe sooner, okay, which is just two months away. So we're excited about a 32.4 overall response rate, heavily pretreated group of patients, many of them that have PARP inhibitors, and this looks pretty exciting. And you can see this swimmer plot with a long duration, so it's not just tumor shrinkage, ORR, it's the duration of response, the number of complete responders, and obviously the side effect profile, which in this case is very acceptable. Most AEs were low-grade. There were some reversible ocular and GI events, and only 7% discontinued treatment due to treatment-emerging adverse events. And then we do a randomized trial to confirm the accelerated approval. Thank you, Katie Moore and a whole host of others. This trial has also finished enrollment. This is also a platinum-resistant recurrent ovarian cancer trial called MARISOL. I'm sure you've heard of it. Primary endpoint is progression-free survival. So hopefully the U.S. will get sort of the SREA indication soon, and shortly thereafter, the rest of the world will get the randomized phase three result in the same setting. We're also doing single-arm study piccolo in platinum-sensitive disease, and that's very exciting. So two antibody drug conjugates. So also, you're familiar with InovA3, another collaboration. I'm sorry that the NGOT number is not on here. NGOT's actually the lead group. Forgive me, Ignace. He did a proof-of-principle study that it's tolerable with tumor-treating fields and electromagnetic radiation frequency that disrupts the microtubules and sort of works, improves, synergizes with paclitaxel to get, here's we go, NGOT-OV50. This study has also completed enrollment. The primary endpoint is overall survival. Again, we hope to have this readout next year and ideally in the clinic next year. PARP inhibitors, we need to overcome resistance. One mechanism is PI3 kinase. You're familiar with that. Panos Konstopoulopoulos, I don't think I said that right, Panos, I'm sorry, Panos as we know him, showed very nice synergy of elaparib and a PI3K medicine, and look at that waterfall plot. Very exciting in BRCA wild-type tumors. So that's led to a study that we're doing now. It's not a GOG study. I think it's an NGOT study. I think so. I apologize for not having the NGOT number there, but it is taking patients who are BRCA wild-type, randomizing them to the combination versus physician's choice chemotherapy. And then Nicoletta Colombo presented at ASCO this idea that we can take this glucocorticoid receptor antagonist and give it to nabpaclitaxel, there's no steroid there, and improve the therapeutic window, okay? So it's an intermittent dose. In this phase three trial, again, I don't know why I don't have the NGOT number there, but it's nabpaclitaxel with or without relacorrelant in platinum-resistant recurrent ovarian cancer. So I'm gonna end here on your wonderful study. Thank you, Isabel. The BOCAE study, we've done very well. In fact, you guys, I think you beat us, not think, I know, and I'm very competitive, as Dr. Slomovic said, but we're gonna roll on the next wave, wave number two. Inovolacib is a PI3K active medicine, and as you know, those can be added with CDK4-6. They can be added with PARP. I showed you Epic-O, and they can also have an antihormonal. In fact, Apeliceb is approved in HR-positive breast cancer. So the PI3 mechanism has multiple pathways that are relevant to rare gynecologic tumors, and then this Giridecitrant is a CERD, and CERD with Abemi and also CDK4-6 is a great opportunity. So thank you for what you're doing for patients, and I know that ovarian cancer is undefeated since, in platinum-resistant disease, since Aurelia, and I know we haven't had any approvals since PALA-1, but stay tuned, because next year is gonna be a great year. In Korea, IGCS. Thank you for having me. Thank you very much, Dr. Monk. Are there any questions for our two excellent speakers? We're running a little bit short of time, but I'll take a burning question. Maybe I can just sort of ask Isabel and Brad about the BOUQUET trial, which is obviously a sort of new paradigm in terms of looking at this. One of the sort of challenges with the BOUQUET trial is looking at the centralized genomic testing. How do we actually make it potentially more accessible in real time with clear testing? You're tough on me. It's a good question, though. Yes, it's a good question. In fact, we have the possibility to test the patient before to be included, and just to remember, the patient can receive four lines of previous treatment before to be included, so the good way to manage is to test the patient at the time of the previous line to be sure that effectively we have the results. But right now, I have to say that I'm quite surprised by the results because at the beginning, I anticipated that it would be very long, but everybody make a lot of effort. All the pathologists who review all the diagnosis works very fastly, and at the end, it's not so bad. We are speaking about resistant population, and until now, I don't have so many issue in term of a patient who are not able to join because they are waiting for the results. Thank you so much, and Brad, in terms of the multiple different types of second-line studies that we're looking at, it strikes me that we need to get more and more patients onto clinical trials. How are we going to do this? Because the sort of, as you said, the standard of care for platinum-resistant disease with conventional single agents is pretty modest, so how do we get more rapid access? You know, that's a critical question. You know, it takes trials. I showed them. It takes investigators. You are them. It takes women with cancer. We have them. It's the fourth component that we struggle with. It's the institutional support. Our institutional people are going and working for pharma. Maybe we don't pay them what they deserve. So I think we have to work with our institutions because we have the studies, the investigators, and the patients. But I think if we could sort of do better at the institutional level, and maybe there are temporary agencies that we can hire. And it takes people that are good at data submission. It takes clinical research coordinators. It takes regulatory. And if your clinical research coordinators are not nurses, it takes nurses and pharmacists. So that's a lot of, it's a village. And so I appreciate you sort of at the Princess Margaret sort of set the standard. And I would ask you, how do you do it? It's challenging, to be honest. I think that we're privileged that we're in a single-payer system and we have a referral system that there's no competition for patients. But to be honest, we're actually also struggling with the same community in terms of trying to make sure that we have the people working in clinical trials because staffing is a major issue. So I think it's a global issue right now. And if I've, you know, I have kids that are in that age group, that's a great career choice. It's a great career choice. You know, get out of college. What do you do with that college degree? You get into the clinical trial arena. It helps patients' lives. There's some reimbursement and some opportunity, not only for satisfaction, but for a salary. With that, thank you so much, both of you. Excellent. Thank you, Dr. Hosa. It's our now second panel in developmental therapy, and I will begin with Dr. Hosa presenting the burning topic about overcoming PARP resistance. Thank you. Okay, thank you so much. I'm going to talk a little bit about overcoming PARP inhibitor resistance. And I think you've heard a lot about PARP inhibitors and resistance developing over the course of the last couple of days, including in some of the presentations that you've just heard. So high-grade serosuvarian cancer is a disease of genomic chaos, and there is a lot of sort of evolutionary pressure in terms of the tumors. Clonal heterogeneity is something that's really, really important to sort of recognize. And the pressures that we put on patients with the treatments actually force further evolution of the tumors and create a more complex heterogeneous tumor atmosphere. There's also significant impact related to the microenvironment. And what this does is that the disease continues to evolve over time, and the evolution of resistance mechanisms is really very much dependent upon the treatments. And I think that this is now becoming more and more important as we sort of really consider the type of treatments that we're offering to patients. We've been looking at platinum-based treatments for the last 50 years, and we're still sort of finessing this in terms of the dose, duration, sequence, cross-resistance, and the possibilities of re-challenge. We've learned a lot by functional algorithms in terms of how we select patients, how we re-challenge patients, and in what settings we actually document that this is platinum-resistant or not. And the definitions of platinum-resistant keep on evolving. To that we've added the fact that we've now added perhaps more chronic treatments with PARP inhibitors. And the chronicity of the exposure is probably going to change the way we sort of really look at how patients are actually tolerating treatments, but also what's happening at the tumor level because of the chronic exposure, particularly related to emergence of resistance mechanisms, and the biomarkers that we can actually follow. It's probably becoming important that we give treatments such as PARP inhibitors and platinum-based treatments that are as effective as possible, as early as possible, and perhaps for a shorter period of time. But what have we learned about clinical features of PARP inhibitor resistance? And I think that we've learned that the activity aligns very much with functional platinum sensitivity. I think at the same time, we've also recognized that there's some more numbers of patients who may actually have intrinsic resistance. It's similar to what we saw with platinum resistance, that there were a group of patients who have a platinum refractory, but we are seeing very few patients who may have genomic scars but are still resistant. I think the consolidation on chemotherapy effect with maintenance therapy has proven to be very, very effective. But the period of time over which we give this and how long we continue really still needs to be further defined. Platinum resistance is associated with poor outcomes with PARP inhibitors, as you've heard. But it's also important to sort of look at PARP inhibitors when we give these to patients, that exposure to a PARP inhibitor does not automatically equate that there's resistance to a PARP inhibitor subsequently. And overlapping resistance mechanisms in PARP inhibitors and the sequence of treatment allow us to sort of really perhaps dig into which generate resistance mechanisms faster. Panos Constantinopoulos showed a really nice curve yesterday, which really looked at trying to identify the chronicity of treatment and how resistance emerges over time, and perhaps trying to sort of really finesse control of the disease with resistance, and perhaps using molecular monitoring mechanisms to try and finesse the time interval that PARP inhibitors are given. I think we could probably add a third curve to this, which is emergence of perhaps other types of complications like MDS-AML with chronic exposure, so that we can actually try and identify which is the optimal time for treatment. So in terms of progression following PARP inhibitors, I think one of the sort of lessons that we've been learning is that clinical trials are really pointing the way to allowing us to understand biologic mechanisms of resistance, and I think we're going to talk about how do we sort of really tackle these. But before we do, it's also important to make sure that if there are patients who have oligometastatic disease, which is truly confined to one or two nodules, that we don't really completely ignore the use of surgery again in some of these patients, and there have been sort of anecdotal reports of many patients progressing on PARP inhibitors with oligometastatic disease that have been successfully resected. So it is important to sort of perhaps look at this a little bit further. In terms of emergence of resistance mechanisms, reversion mutations, as you know, have been really the most described and most frequent resistance mechanism that people have identified, and clonal heterogeneity really leads to sort of multiple parallel reversion mutations sometimes in the same patients, and this really does predict for extreme resistance to PARP inhibitors, but also probably to other types of therapies as well. And there's a really nice meta-analysis that was done recently that perhaps shows that reversion mutations may be present in as many as about a third of patients, and there seem to be perhaps hotspots related to the genomic scar related to BRCA1 or BRCA2 that may also predispose to reversion mutations, and some of the things that we're learning are that there are sort of specific aspects related to microhomology-mediated end joining, which could be potentially an area where there could be targeted therapies to perhaps reduce the risk of reversion mutations. One of the other intriguing findings was that we recently just reported the results related to ARIL4, and one of the sort of additional sort of facets that were identified from ARIL4 is that in patients who had reversion mutations, when you actually look at the PFS curves on the right-hand side to how they responded, generally they respond poorly to treatment, but perhaps they respond better with chemotherapy, which was weekly paclitaxel, compared to PARP or other types of therapy. And reversion mutations seem to be sort of significant in terms of driving some of the sort of mechanisms of resistance, and as you saw, patients who had chemotherapy did better in platinum-resistant disease compared to patients who were treated with a RUCAPRIP. But one of the sort of other intriguing sort of findings is that the frequency of reversion mutations seemed to diminish in patients where we had sequential specimens over time when they were treated with weekly paclitaxel, so again, it perhaps indicates that there may be a better outcome with patients who have reversion mutations if they're treated with weekly paclitaxel. And what was intriguing is that when patients who were on the weekly paclitaxel arm were subsequently treated with RUCAPRIP, there was an improvement in terms of their time on RUCAPRIP after crossover. Another study that we sort of reported a while ago when we were trying to look at how do you tackle resistance post-PARP was the evolved study where we added in siderinib in addition to PARP inhibitor, and this busy slide really illustrated that on the left-hand side were the sort of de novo patient assessment in terms of resistance mechanisms from their initial biopsy at the diagnosis. And following exposure to platinum and PARP on the right-hand side, a biopsy looked at the different types of resistance mechanisms, and as you can see, there are multiple different types of resistance mechanisms that developed in patients and sometimes multiple mechanisms in the same patient after exposure to PARP inhibitors. Transmission mutations was one of them, but there are many other mechanisms of resistance that developed. And this summarizes perhaps some of the different mechanisms that are actually being seen and how do we tackle them. There are multiple mechanisms of resistance which actually show that restoration of homologous recombination, replication for protection, signal transduction pathway adaptation, or upregulation of drug efflux all lead to increasing resistance developing post-PARP inhibitors. And over time, as clones of cells continue to survive with developing resistance mechanism and the sort of clones that are sensitive continue to go down, what is left is a much more resistant tumor which then needs really other types of mechanisms to try and overcome that resistance. And what are the sort of different types of therapeutic strategies that we can use? I mean, first of all, you can potentially look to see if patients are not truly resistant, whether there are any patients who would benefit from PARP rechallenge, and I'll come back to that. The other way of perhaps overcoming resistance is to try and see if we can actually resensitize tumors by overcoming resistance. And you've heard about some of the potential studies going on from Brad. And then bypassing resistance by really looking at non-crust-resistant agents, and again, Brad and Isabel talked about some of the sort of newer targeted agents which are really completely different in terms of their targeting, which do not really depend on PARP inhibitors. A study that really informs us about perhaps rechallenge with PARP is the OREO study run by the Gineco group. And again, what this showed is that if you carefully identify patients who had a good response to PARP inhibitors and rechallenge them, potentially there may be some patients that can benefit again. The overall impact in terms of the clinical impact of rechallenge with a PARP inhibitor looked quite modest, but there is a tail on the curve, and there are some patients who had been previously exposed to PARP inhibitors and had stopped, who may perhaps benefit from rechallenge. But we need to be able to identify these patients carefully in order to be able to see if we can really rechallenge patients. There are many different mechanisms in terms of overcoming PARP resistance, and I sort of really direct you to this recent article in Nature Reviews, which really summarizes the different mechanisms, particularly looking at immunotherapy, targeting the C-gas-sting pathway, looking at introducing tyrosine kinase inhibitors, as well as looking at inhibition of parolation by interfering with NAD synthesis. Also looking at the restoration of the replication fork, and looking at Pol-Q inhibitors. And Shannon may talk about some of these, but all of these are different mechanisms to really overcome resistance. But it is important to sort of really identify which patients we can truly overcome or help in terms of resistance, and which patients may in fact do very, very poorly. And in our evolved study, there were some patients where despite adding in a tyrosine kinase inhibitor, the potential impact was very, very low, particularly if they had ABCB mutations or patients who had reversion mutations who did very poorly with 3-challenge. We also really looked at trying to look at restoration of replication fork stability using a V1 inhibitor. And in a very resistant refractory patient population, there seemed to be a significant synergy between adding in gemcitabine plus a V1 inhibitor in terms of patients with platinum as well as majority with PARP resistance as well. And Shannon, again, presented these results of the effort trial looking at V1 inhibition just recently at ASCO. And this study really looked at V1 inhibitor Adavacet by itself, as well as Adavacetib and Olabrib as a combination. And there seemed to be a significant synergy when you give the two agents together with a significant proportion of patients receiving clinical benefit as well as objective clinical response. So I think that there are potential mechanisms that we can exploit in terms of trying to improve on potential synergies with PARP inhibitors. Finally, perhaps these agents in terms of looking at bypassing resistance altogether, I think Brad dealt with this really nicely with antibody drug conjugates, particularly looking at folate receptor targeting in the SORAYA trial. But I should just point out that in the SORAYA trial, a significant proportion of patients had previously already been exposed to PARP inhibitors. And the response rate, as he illustrated, looked really quite encouraging. So in patients who had had PARP exposure previously, about 38% of patients responded very well to folate receptor targeting with Mervitaximab. So again, this really does illustrate the potential to bypass resistance and look at other targeted mechanisms. So I'd like to sort of summarize by saying that there are multiple resistance mechanisms to PARP inhibitors, and the sequence of platinum and PARP is important. These resistance mechanisms overlap, but the duration of therapy also probably likely influences emergence of resistance and early effective intervention for a limited duration of time. Perhaps maybe a better strategy looking long term. And these results from SOLO1 as well as PRIMA are really encouraging in terms of early introduction of treatments. I think we do need to understand emergence of resistance, and this will allow us to design studies to bypass, overcome, or re-challenge patients appropriately. But in all of these settings, I think it's really important to integrate predictive biomarkers to track and predict failure. And clearly, we need to identify ways of reducing emergence of resistance and if there are sort of effective therapies to help us overcome or bypass it. So thank you very much. I always like to sort of end my talks by acknowledging the patients who participate in our studies. Thank you. Thank you very much, Dr. Hauser. I'm pleased to invite Dr. Shannon Westin to speak now about delivering new pathway in ovarian cancer. Thank you. And really great talks. You guys did an awesome job putting this together. So it's my pleasure to talk a little bit about leveraging new pathways for ovarian cancer. And I hope that this talk will be complementary to what we just heard from Drs. Monk and Oza because we really are needing to push the envelope once we get beyond the first-line treatment. And these are my disclosures. So first, I'm going to start with anti-angiogenics. You've heard quite a bit about that during this session and also throughout the conference. We know that bevacizumab is an excellent option for our patients, especially for patients for whom platinum is not an option. Of course, that was based on the Aurelia study, which demonstrated improved progression-free survival when we added bevacizumab to standard-of-care chemotherapy. And importantly, because whenever we're adding something to our patients, we want to make sure we're not negatively impacting their quality of life. And we didn't see that. Patients actually felt better. They had disease control and they had less symptoms. And so this is certainly an excellent option. However, we know that patients will develop or even have innate resistance to these agents. And so what do we do if we have a patient that has had progressive disease on an anti-angiogenic? Well, one of the mechanisms of action that we've seen in our own work is the activation of the NOTCH pathway. So activation of the NOTCH pathway can lead to resistance to anti-angiogenic agents. So we initially did some studies looking at a DLL-4 antibody inhibitor. Now DLL-4 is a part of the NOTCH pathway. And as a single agent, it did quite well. It seemed to overcome resistance to bevacizumab in a number of xenograft models. However, we wanted to try to push the envelope. And so we were able to develop a dual inhibitor that targets not only DLL-4, but also VEGF. And with that dual inhibitor, we've seen both as a single agent as well as in combination that this is quite active. You can see on the left the results of the single agent trial in phase one. And when we teased out the patient population that had epithelial ovarian cancer, it was about a 25% response rate. And the majority of those patients in that first trial had had prior bevacizumab. This went on to be combined with weekly paclitaxel and platinum-resistant ovarian cancer, and again had quite a high response rate despite median prior therapies that approached five or six. And you can see that in patients with prior bevacizumab, it was about a 33% response rate. And for patients with no prior bevacizumab, we saw higher rates of response of 64%. But still a really nice-looking waterfall curve, and this agent is under evaluation now in a phase three trial, which randomizes between weekly paclitaxel with or without the nabasiximab. Now, you've heard quite a bit about antibody drug conjugates, so I'll highlight some of it and then move on to some of the more novel things that I'm going to discuss. But first, I just want to make sure we're all on the same page about what these agents can bring to the table. Because they have these three specific parts, we can really start to target and utilize precision medicine for these patients by assessing what they express. So each antibody drug conjugate is going to have that antibody that's quite specific to an antigen that's hopefully preferentially expressed mostly on tumor cells and then link it to a very, very high-dose payload that normally could not be given on IV. This is a table that I stole from the fabulous Dr. Ray Kocard who put together a really nice summary of all the antibody drug conjugates that are in exploration in epithelial ovarian cancer. And you've heard a little bit about the first two, the folate receptor alpha as well as NAPI2B, and I am going to reiterate those because as Dr. Eskander said yesterday, that repetition helps us learn. But I also want to call your attention to a few others that are quite early in the assessment, including a TROP2 antibody, which is showing quite a bit of activity. And we know that endometrial cancer and ovarian cancer both express quite a bit of TROP2, and so this might be an interesting one. There's a MUX16, which is also known as CA125. That antibody drug conjugate will be of interest. And then of course, tissue factor, which has predominantly been utilized in cervical cancer. So first, just a little discussion about MIRV atuximab. As you know, it is targeting folate receptor alpha. And I thought it was important to kind of just mention the history because it really does inform how we're selecting patients for trials, how we're doing the molecular testing to really guide our patient population. Now MIRV was initially assessed in early phase trials, showed great activity, and that led to Forward 1, again, in platinum-resistant ovarian cancer. And they did try to tease out medium or high folate receptor alpha positive expression. They were randomized 2 to 1 to the MIRV atuximab versus investigators' choice chemotherapy. And unfortunately, the top-line results, the intention to treat results, were negative. And what we saw was there was no improvement in progression-free survival for MIRV, even really when they teased out the folate receptor alpha high. But it turned out it was really mostly based on the scoring system that was used. They had moved from a PS2 scoring system to a different scoring system. And when they went back to their initial folate receptor alpha high scoring system, they did see a difference. And it's just a testament to the commitment of these investigators and the industry partner that they moved forward and are now doing these trials or did these trials in the right population. And you did hear about these from Drs. Monk and Oza, so I won't belabor the point. But the fact is, is when we appropriately select the patients, MIRV atuximab works. And this is potentially going to be something we can be offering our patients over the next few years. Oupafetamab rilzedotin, or UPRI, which I'm going to call it, is another antibody drug conjugate. This one targets NAPI-2b. Now, this is a sodium-dependent phosphate transporter. And it's expressed in about two-thirds of ovarian cancers. We're still trying to determine how much folate receptor alpha expression overlaps with NAPI-2b because, of course, that'll be very informative for our patient population and our choices. But it does seem to be another great option for a large proportion of our patients. In the early ovarian expansion of the phase 1b, they really worked out the right dosing strategy. And in the ovarian cancer cohort, it wasn't a very heavily pretreated group, about one to three prior lines. And they saw really nice efficacy, especially, actually, in the lower dose group, so teasing out the best dosing strategy for this patient population. It was about a 44% response in that NAPI-2b high group, so a molecularly chosen group that got the lower dose. And that, as you saw, has led to uplift as well as up next. So let's get into some other kind of interesting targets. We heard a little bit about an ongoing trial with glucocorticoid inhibitor. I thought I'd get into the details of the pathway a little bit more. So what we know is that cortisol can suppress immune activation and tumor cell apoptosis by activating the glucocorticoid receptor. And then we see tumor progression and induced chemotherapy resistance. So we see high glucocorticoid receptor expressed in a lot of solid tumors, including ovarian cancer. And so there was interest in utilizing this selective glucocorticoid receptor modulator to enhance or restore chemotherapy sensitivity. And as you saw, when we looked at three different arms, including an intermittent relacorlant versus a continuous relacorlant, it actually recapitulated what they had seen in the preclinical data that the intermittent dosing strategy was better. And they did combine it with the NAB-paclitaxel, so you didn't have to use a steroid premed, and found that it did improve progression-free survival, which is now leading to a randomized phase three. Now, AVB500 is an AXL inhibitor. So AXL is highly, highly expressed in metastatic ovarian cancer. It binds to GAS6, which is a sole ligand. And what we see is, I'm sorry, AVB500 is a high-affinity AXL decoy. So it binds actually more tightly to GAS6 than the wild-type AXL that's in the patient population. So in a phase 1b, combined with paclitaxel, it was quite well-tolerated, and response rates were 41%. So we do have an ongoing phase 3 study combining weekly paclitaxel with or without AVB500. Tumor treating fields are super cool. So this is, it's not necessarily a pathway, but still something that I think, a new modality that is very exciting. This is basically where we see electrical fields that induce forces on charged tubulin proteins, and this basically acts almost like a taxane, where it disrupts the formation of the mitotic spindle. It is approved in GBM and mesothelioma. And in an early phase trial, we saw a median PFS of 8.9 when we combined weekly paclitaxel with tumor treating fields. Response rate was 25%, and that led to the recently completed phase 3 trial, which is super exciting. Now P53. This is like our holy grail, right? We know that P53 is highly overexpressed in, or highly mutated in high-grade serous ovarian cancer, and so it's long been something we want to target. We haven't gotten to target all of P53, but this new drug that targets the Y220C hotspot mutation is certainly very exciting. We had this open in our institution in phase 1. It's only present in about 1% of solid tumors, but we do see it in patients with ovarian cancer. You can see about 2.9% will express it. And the drug works. So this is PC14586. It's a selective Y220C first-in-class P53 reactivator. It hits this key hotspot and led to a 32% response rate. And again, this was a highly pretreated phase 1 patient population with medium prior therapies somewhere around 5. So this is a drug that kind of gets at that idea that we can utilize precision medicine and select out a rare tumor to treat a patient. So finally, I'm going to end with adoptive cell therapies. I think this is a very savvy audience, but this is where we're collecting T cells from a patient or growing them in a lab, putting them back into the patient where we can increase the number and also increase the strength and the effectiveness of a particular T cell. Now we sometimes do need a specific HLA subtype, often O2O1. And this particular adoptive cell therapy that's moving forward in ovarian cancer is no different. So ADPA2M4CD8, say that three times fast, this is a next-generation T cell therapy that targets MAGEA4. Similar to what we see with UPRI, similar to what we see with Mervituximab, this is highly expressed in ovarian cancers and is a very relevant agent. And what we see when we looked at a phase 1 trial that was just recently presented at ESMO, a high proportion of patients with response rates, so 28%. And when we teased out that population, it really was predominantly patients with ovarian cancer that had the best benefit. So this is moving into a phase 3 study, the Surpass 3 study, that you all will be able to open up in your clinic. So it was a whirlwind tour with a very short period of time, but hopefully I've convinced you that we've got a lot of really exciting novel therapies that are coming down, including new technologies as well. But we really need to be thoughtful about how we design these trials to ensure we're assessing predictors of response and resistance to kind of push those rational combinations that Dr. Oza mentioned and improve outcomes for our patients. Thank you so much for your attention. Thank you so much, Dr. Westin. I will be, unfortunately, the bad girl, and I will do not give you enough time for the question. I will move to the next panel. Sorry for that. Thanks so much. Our next panel is current level of evidence around HIPAC. And we'll begin with Dr. Kathleen Moore presenting HIPAC, a critical appraisal of promises and challenges. And while she's coming up, as Isabel mentioned, I think because of time, unfortunately, we'll not be able to do Q&A. But I would encourage everyone to engage the speakers after the presentations are done. I was so enthralled by Dr. Westin, I just kept sitting over there and didn't realize I was next. Whew. Okay. That was my exercise for the day. Well, I'm going to give us an overview of HIPAC, really the promises of HIPAC in ovarian cancer. Oops. Here are my disclosures. And what I'm going to do is really talk on why heated regional therapy is logical for treatment of advanced ovarian cancer. So we'll discuss in general the use of regional therapy, review some of what's known about molecular and immunologic rationale, and summarize the available clinical data. And then I'll turn it over to Dr. Herzog who will discuss some of the unanswered questions and challenges regarding HIPAC in the next talk. So let's get ready to warm up to HIPAC, or we may focus on keeping it cool. So regional therapy in ovarian cancer makes sense. It should work. Ovarian cancer tends to be a peritoneal surface malignancy. We know that it sheds directly from the tumor into the peritoneal cavity where it can establish implants and vascular networks, et cetera. So we should be able to utilize this pattern of spread to our benefit for surgery and for regional therapies. And a lot has been done with regional therapies outside of gynecologic cancers. We all know this. Colorectal cancers kind of have led the field. And because of this, there's been a lot of standardization around the language we use when we talk about the cytoreduction that happens before we use these regional therapies. And I just give you two examples here. And this is really important because as we do studies, we need standard language. We can't just say optimal. We need standard language for how much tumor was really left, ideally none, because any regional delivery of chemotherapy will penetrate only 5 to 8 millimeters. So really, cytoreductive surgery down to really no gross residual, 2 to 5 millimeters is crucial for this to work. And why does it work? So it's rational. You have distribution of chemotherapy in the belly. So the literature talks about this bidirectional model for IV and IP therapy. So the intraperitoneal or belly chemotherapy reaches the outer layer of the tumor, which is ideally microscopic, and can penetrate there with very high concentrations due to direct exposure. And then you give intravenous chemotherapy, which can get to the inner core of the tumor due to microcirculation. And then there's a barrier that prevents much of the intraperitoneal chemotherapy from crossing into the vascular network and getting us too high in terms of drug exposures. And so regional chemotherapy is not new. We have done normothermic intraperitoneal chemotherapy. These are the five big studies, and you know them well. The top three, actually, four of the five were positive for progression-free survival. First of all, we can say that. And we'll talk about 252 in a moment. But the first three, GOG 104, 114, and 172, led to this clinical alert in 2006. There were meta-analyses done, which you can see at the top, for progression-free and overall survival that really favored use of intraperitoneal normothermic chemotherapy. And it led to this clinical alert, which really recommended use of intraperitoneal chemotherapy for patients post, quote, unquote, optimal cytoreduction. But despite that, the uptake was really minimal. And you can see there was a little blip in 2007 and 2008 in the yellow and gray bars for utilization of intraperitoneal chemotherapy, normothermic, in the United States. And then it's gradually tailed off, really, by 2012. And in Europe, it was really never endorsed. And I took this from Dr. Dubois' discussion in 2018. It was really relegated to clinical trials because there wasn't felt to be enough demonstrated benefit to justify standard of care use. And I think we've kind of come to agree with that. We saw 252 reported by Dr. Joan Walker at SGO in 2022 for overall survival. And I'm showing you the group that we should be talking about here for HIPAC, which is those who have microscopic residual disease after primary cytoreduction. And you can see the three arms here, which is IV carboplatin, and then the two IP experimental arms, and they all clearly overlay one another. And you may say, well, there wasn't a difference because of the bevacizumab, and the bevacizumab equalized this. Well, let's overlay GUG172. And I borrowed this slide from Dr. Michael Bookman, who gave a master class discussion of this. And what you can see is that the IP arm of 172 completely overlays 252. It's cross-study comparisons. I understand that. But it really doesn't look like bevacizumab is the reason this was all unequal. It looks like the IV cisplatin arm was inferior, which is what we've always assumed from GUG172. We don't think the OS was improved with the addition of bevacizumab. Contemporary chemotherapy with IV carboplatin really does appear similar to IP. So this was kind of the nail in the coffin for normothermic regional delivery, in the U.S. at least. And so why do we think heat can fix this? That doesn't really make sense. Well, maybe it does. The idea of hypothermia or hyperthermia, pardon me, being anti-tumorigenic has been postulated since 1890s. It's really an old concept. And we have actually preclinical data, and this is a wonderful review I would encourage you to take a look at. I've just pulled some tables from it. We actually have cell line data that shows that there's cell lines who are susceptible when to get to 42 degrees to cell death. But the interesting thing, and I'll come back to this point, is that it's not all cell lines. Some cell lines require much warmer temperatures and for much more prolonged periods of time to enact cell killing. So while this does suggest that hyperthermia leads to cell death, it doesn't tell us what temperature and for how long. More preclinical data has been done that also suggests that there's synergy generated by hyperthermia, that when you expose these cell lines, especially those that are cisplatin resistant to heat, you do regain sensitivity to platinum, which is interesting, and you show synergy. And so a lot of this is supportive of moving this into clinical practice. This is another great review article that came out last year by Thandelliger et al. that kind of goes through all the hypotheses as to why this works. Causes apoptosis, you have heat shock protein activation, inhibits DNA repair, T cell immunogenic infiltration, increased blood flow, et cetera. So let's go through this. Hyperthermia may degrade BRCA. That's a theory. And so it's not been studied in ovarian cancer, but maybe that turns it into a BRCA-like phenotype. But it also can upregulate heat shock protein 90, which protects BRCA, and can also upregulate heat shock protein 70, which can lead to pro-tumorogenic interactions with client proteins that are favorable to the tumor. So it's really unclear how this, if this is a favorable or an unfavorable interaction. The other theory is that increased perfusion will bring more tumor-infiltrating lymphocytes. Hyperthermia can upregulate ICAM-1, L-selectin integrins, which should make, you have more natural killer cells, et cetera. But this is a theory that has not yet been proven. Heat can cause vasodilation, which may improve blood flow and anti-cancer agents in immune cells, bring them more into the tumor. All good theories and hypothesis-generating, but unproven. We do have a little bit of data, and this is courtesy of Oliver Zivanovic, who actually did biopsies and looked in the surgical specimen pre-HIPEC and then post-heated chemotherapy, and did show improved, increased cisplatin DNA attic formation, which lends some credibility to the idea that we're inducing DNA damage response inhibition, which would be favorable to chemotherapy. So this does support that hypothesis, but still more data is needed. And as I alluded to, the impact of heat appears to be somewhat dose-dependent. And what you get at 37 or 38 degrees with vasodilation, maybe some impaired DNA repair, you may not get cell response until 40, increased drug uptake around 42, but direct tumor cell lethality requires much higher temperatures. And some cells require different temperatures than others, and we've never really truly figured out what the right temperature is for a high-grade serous versus an endometrioid versus a clear cell, and does it matter? Usually no one knows, and we should know this to optimize it for our patients. What's the ideal chemotherapy for HIPEC? Mainly this is the list. A lot has been used outside of gynecologic cancers, but cisplatin is our main asset with about a 20-fold increase in peritoneal versus plasma ratio, and you can see that outlined here in the figure on the right-hand side in orange as compared to plasma levels. You do get very high concentration of cisplatin, intraperitoneal, and also carboplatin. This is actually data from Dr. Zivanovic's clinical trial where they obtained specimens at the time of cytoreduction and definitely showed equivalent increase in carboplatin concentration as compared to plasma with intraperitoneal heated carboplatin. These are our two main agents. What about the clinical use? Well, there's two techniques that kind of limit cross-study interpretation, but we have this open technique which is an advantage because you can use your hands to distribute the chemotherapy. But the disadvantages are you can expose your whole OR to chemotherapy and spill it, and it's hard to maintain heat. You have accelerated heat loss. There's the closed technique where you have minimal heat loss. It's easier to get the intraperitoneal temperature to 42 degrees. You reduce spillage, maybe improve drug penetration, but you don't really know that you're getting it all over the belly, and there may be a higher concentration of drug in the blood. But this all moved into clinic. There was a lot of retrospective data published mainly in the recurrent space, which all looked pretty good and justified moving the field forward. There were meta-analyses done both in the primary, and those are the As of these meta-analyses, and then recurrent in the Bs for both PFS and OS, and you can see that all of these favored intraperitoneal chemotherapy or favored HIPEC. So moved into clinic in the recurrent setting, and you can see the studies here. Only two of these are prospective, Spiliotos and Zivanovic. Spiliotos is a larger randomized phase three, but both done in two different settings and was positive. Zivanovic was a smaller study, but only done in the recurrent setting and was definitively negative, so very inconsistent data here. Nonetheless, it moved into frontline with a small phase two, which showed an advantage. And then we had the LIM and Vandriel studies that you all know well, again, some inconsistencies. LIM was negative. Vendryl was very positive for both PFS and OS. And I put this up here just to remind you of what that study looked like and the PFS and OS. But there are problems and questions of this. And again, I'm going to put up Dr. Dubois' slide. Thank you, Dr. Dubois, for this excellent slide from the discussion in 18, just to point out. And again, it's cross-study comparisons. But the control arm in the Vendryl study, the study took nine years to accrue. And the control arm is really poor, 11 months versus 34 months for OS as compared to all of these other studies. So there's some kind of selection here that was potentially confounding the results. And the other thing we have to consider about is that now we have, in terms of benefiting HIPEC, how do we interpret that study in relation to all the changes in standard of care, like PARP inhibitors? Will it be additive or synergistic or no effect when maintenance is added? So I've shown you the Vendryl PFS on the left. And underneath is PRIMA, which are similar groups, neoadjuvant, and they're different scales, which is why they look different. But if you look at the box, you can see that the Nurapper group is 67% neoadjuvant versus 100% in Vendryl, so a little bit different. The progression-free survival in your study arm is 13.8 versus 14.2 months. Cross trials consideration, yes, but still, do you need HIPEC if you have PARP? Do they add together? Do you need one or the other? We don't really know. And the standard of care has changed, at least for now. So these are the ongoing studies. In HIPEC, there's a lot of activity. The green are completed and reported. Yellow is repeated, completed, and not reported. All these other ones are ongoing. And we have a study that's just activating GOG3068, led by Oliver Zvanovec and Leslie Randall, that's trying to account for a lot of the things that we're missing in the Vendryl study. It's stratified by HRD. Everyone gets maintenance chemotherapy, and we're incorporating some translational work in this as well, which hopefully will add some to the literature on whether or not HIPEC should be the standard of care in this patient population or not. So in conclusion, there are many proposed mechanisms by which heat may augment responsiveness to chemotherapy, but most are unproven at this point. So this is a big need. The current trial, Dr. Vendryl's side, supports consideration of HIPEC but leaves a lot of unanswered questions. Was the population overselected? It wasn't stratified. Is that important? If you had stratified, would the outcomes have been different? And how does this maintenance negate or enhance HIPEC? So ongoing studies will hopefully add more information to either or all of these questions. So at this point, still the question is, are you ready to light a fire under HIPEC use globally or cool it down? So Dr. Herzog, I'm going to turn it over to you. Well, we're 53 minutes behind, and I know the cocktails start almost now, so we have one debate to follow. So I'm going to really skip through here pretty quickly and try to keep it moving. So I'm here to look at the unanswered questions, and I think Dr. Moore actually already framed some of the pertinent questions that are out there. These are my disclosures here, nothing directly involved in what we're talking about here. So as Dr. Moore said, we're looking at high-tech, and I think that for me, you know, we've seen a lot of different platforms of what to do after standard chemotherapy, and we've spent really the better part of the last almost two decades trying to figure out maintenance therapy, dose dense, incorporation of biologics, and the whole conversation about intraperitoneal chemotherapy, and yet we still don't have a resolution of high-tech. If you go back in the literature, you can see pre-clinically it's been talked about for a long time, all the way back into the 1800s, but importantly as a therapeutic modality since the early 1970s, so approximately 50 years we've been looking at this, which makes you wonder why we haven't figured it out yet, and I think it's a lot of polarity. So a lot of people go into this with a fair amount of bias, if you will. There's some people who report this as a potential cure. Other people think it's almost a placebo. There's the hope and hoax. In fact, I wrote an editorial on high-tech probably over 10 years ago that was titled High-Tech Hope or Hoax and went over all the data and where we stood with that. I'm not going to waste time reading all those to you in terms of standard of care or experimental at best and so forth. The IP trials we just covered, so I'm not going to talk about that. So the only thing I will say is this whole scenario has followed a similar course to what we've seen with secondary cytoreduction, but to a smaller volume, and that has been that there's a number of trials out there that were single arm that have taken highly selected patients and said, look, this seems to work compared to historical controls, but there was a lot of selection bias that went into those. So I think moving forward, we can only look at randomized trials. They're the only trials that really matter in this area anymore, much like we saw in secondary cytoreduction where we got much more information by doing randomized trials. We've seen the Vandriel data, so I'm not going to go over that again. This is the LIM data, so we've seen positive data that came out with almost a 12-month advantage in overall survival, and yet here's the LIM data that came out this year that shows in the total group there that there's really no significant difference in progression for your overall survival, and you can see the p-value 0.43 and 0.52. However, if you break those groups out by neoadjuvant versus primary cytoreductive surgery, you can see that there is an advantage in the neoadjuvant group, just like we saw in the Vandriel data. So that gives us hope then, while there isn't an advantage in the primary cytoreductive group, that there could be still this signal in the neoadjuvant group, and it was under power to really show anything beyond that, but if you can see there, the hazard ratios were pretty impressive in that group, with the PFS hazard ratio of 0.6 and 0.53 for overall survival, so significant advantage again in the neoadjuvant, and it's interesting how little it did in the primary cytoreductive group, the top two curves there on the slide. Well one good advantage of getting your slides in late is that you can even incorporate things that were done yesterday, so this was in the plenary session yesterday, and I'm able to show you this data. So this was Dr. Lee from the Korean group looking at HIPEC following interval cytoreductive surgery, and these patients went on to, some of them got maintenance and so forth, and we'll talk a little bit about that, but this is the primary endpoint of this trial, was actually progression-free survival, and you can see here positive data with good spread in the medians and a hazard ratio of 0.61. Here's the overall survival data, again very impressive in terms of the hazard ratio that's seen here, and so very good data. The bad thing about not having much time to get your slides in is that you couldn't remake this table, so what I wanted to show on this is if you look at the maintenance, the point estimate is very far to the left favoring HIPEC, however there was an imbalance in maintenance therapy with more maintenance being given on the HIPEC arm, so that really conflates the results there for us and being able to try to figure that out in terms of what that means. So really where are we? And Dr. Moore said it well, and I didn't know if she was going to bring this in, but you have to look at this relatively old modality and put it into the context of our modern therapeutic landscape, and so you have to think about would these other agents in these new approvals be able to do the same thing as HIPEC or more? What are some of the unanswered questions with HIPEC, and this is really the crux of the thing, and I'll try to do this in just a few slides, but I wanted to acknowledge Neil Sood who's done quite a bit of work in all our basic sciences, and you saw his excellent talk today, but one thing that was very interesting is we dug this up, and Katie Moore, thank you for alerting me to this, was the connective tissue growth factor concept that really promotes hyperthermic resistance. So the expression of this and other cytokines can actually change the response to hyperthermia. So not all cell lines respond, as Dr. Moore said, and this was a really good paper that looked at this, so Dr. Hatakeyama from Japan, and with Neil Sood is the senior author on this, came out in 2016. So question then is could you possibly target this CTGF, thereby sensitizing the cells, and that indeed was done in orthoptic mice models, and so here you're looking at the variability, and if you look in the lower, you can see the differences there. In the abscissa, you have the temperature, and then you have the cell viability on the ordinate there, and you can see the different responses to hyperthermia with different cell lines, and so you have hyperthermic sensitive, which are in the blue where the tumors shrink, and then in the red in the lower panel there, you see the ones that are resistant, and here you can see on the top there on the left is the volcano plot that looks at the genes that are up and down regulated in cells that are hyperthermic resistant versus sensitive, and then in the B panel there with the red and the green, we're looking at the heat map of expression of various of the top 30 genes that are up and down regulated in cells that are hyperthermic resistant, and the deeper the color, the more resistant or sensitive they are, or more up-regulated or down-regulated they are rather, and on the bottom panel there on the left, you see a network map of these genes, and I'll let Anil in detail explain this at about 7.30 tonight, he'll give you more details on that. So this leaves us with the questions of really where we are with this. So in the Vandriel data, is this going to be reproducible, or is this a type 1 error, and now we have other data that reinforces that, so I think we feel good about this data. Relatively small phase 3 trial compared to what we normally see, so you always do worry about a type 1 error, but I think it was a well-done trial. There are some imbalances, there were some differences in the rates of ostomies, and so there's some criticisms within that trial, but nonetheless, the other data that we're seeing is encouraging. It's the clinical effect that we're seeing just due to higher platinum exposure. So I think a lot of people, much like we saw with the IP, would want to know if this is a dose density or intensity problem. And so we haven't really controlled for that very well, and so we've certainly tried to do that with some of the IP trials, but that's been difficult. But I think there's something to think about there. And then, what molecular subtypes should receive HIPEC, and this is an area that's been understudied to date, so many of these trials were commissioned before we understood the importance of BRCA mutations, and specifically even HRD, and so who should be included in these trials, and how do we manage that? What's the role of HIPEC in primary cytoreduction? It doesn't appear to be very beneficial from some of the randomized data that we're seeing. Why is that? We need to better understand that, and can the gains that we're seeing in recurrence-free survival in OS be reproduced by modern therapeutics, and why does HIPEC appear to be tumor-specific? So why are we seeing now negative colon phase III data with very robust trials? So we need to think about that. Certainly ovarian cancer could be different, but a lot of people predicted HIPEC would be highly effective in colon cancer as well, and that has not been the case. We've also noted that actual achievement of the temperature in the tumors that is the target has often been not able to be achieved, so we need better understanding of what the temperature needs to be in these tumors. And I think, as a shout-out to the basic science, and that's why I showed a couple of slides on that, is we really have a rudimentary understanding of what genes are responsible and what biomarkers might predict response or non-response to hyperthermia, and it may be even that hyperthermia actually could be harmful in some patients, so we have to really figure that out as we move forward. Safety profile has varied across different trials, even some of the larger trials, so we need to look at that, but it's been more consistent within the neoadjuvant cohorts. So this is the trial that I think will answer that. The reason being that I think that you need to do several things. One, you need to have a contemporary landscape, which we do here, and you need to control for maintenance, and that will be controlled for in this trial, so GOG 368. And so we'll be able to take care of those two criticisms that we've seen in other trials. We also have good stratification factors that I think will be helpful, and I think this is a trial that will be very helpful in terms of flushing out some of these criticisms of what we've seen and eliminate some of the questions that we currently have about HIPEC, so please enroll in this trial. We're just getting going, and we're looking for sites, and we're very excited about it, so just contact any of us. So in conclusion, we still have a lot of unanswered questions. We need these adequately powered stratified phase three with control of maintenance that I just talked about. Continued translational work needs to be done, and then we need to look at that in comparison to the modern landscape of therapeutics that we have available. Prime time for HIPEC may be nearly here, so please enroll to 3068, and let's make it happen. Thank you. Thanks so much, Tom. At this point, let's go ahead and move on to the final debate, and Tom, will you introduce the speakers? Thank you, Andy. Thank you, Anil. The topic of the last debate is going to be secondary cytoreduction by Dr. Andreas Dubois, who's going to take the pro, and Dr. Rob Coleman, who will take the con. Dr. Dubois, we're going to have you begin. This is like the rematch, at least number three, of Muhammad Ali versus Joe Frazier. You folks have done this around the world, so we're excited to hear the latest rendition. Thank you, Tom. It's a little bit like this. When I got the invitation to fight again against Rob in his own country, in his own Congress, when he's the president of this meeting, I thought it's a cheap way to search for a victim. Anyhow, I will give my best. So these are my disclosures. So surgery is a mainstay of therapy. We heard a lot about all new drugs and all new approaches, but still surgery has something to add to all that in ovarian cancer. When you look at early disease, when you look at the old studies of action and ICANN1, you can see that 60% of the patients are cured only by surgery, and only very few of them were comprehensively staged, and the chemotherapy added about 10% to 15%. In advanced disease, it's prolonged survival, and it increases the rate of long-term survival. So why shouldn't it work in recurrent disease as well? So anyhow, but this is belief, and Tom just said it. There's a lot of belief about a lot of therapies, but what we need is data, and that was the reason why we started the desktop series. In those days, surgery in recurrent ovarian cancer had no evidence, and so we thought it would not be fair to expose patients to a toxic therapy without knowing whether they have at least a chance to benefit. So we started with Desktop 1 to identify the population which may have a benefit of surgery, and we found that only patients with a complete resection were associated with a better outcome. That was confirmed later in other studies as well. And then we tried to identify these patients, and we built a hypothesis for a score with data available before we go into the OR room, and this is clinical data. So we had imaging that ruled out non-resectable disease, and then we had the clinical data with performance status, residual disease after frontline surgery, and ascites, and we built a score of it, and then we put that to a prospective multinational, multi-continental trial and evaluated that score, and that was Desktop 2, and then we could confirm, we could validate the score with a statistical assumption that more than two or three patients with a positive score later will have a complete resection with 95% probability, and we met this criteria. So we now had a score to select patients which have a chance, have a high chance of having a complete resection, and these were the patients where we believe that this is a population where we should test surgery for recurrent ovarian cancer because they have a fair chance that this cohort is the best cohort for this. And so we started Desktop 3. In the meantime, Rob has already started GOG 213, which was a twin study looking on bath and on surgery, and later on, the Chinese group, together with Ronyu Zhang, split off Desktop. They started with Desktop, but then they worked further on their selection criteria and found their Tian score, and as you can see on the author list, this was based on our retrospective data and other data, and so they split off Desktop and found their own study, which was SOC 1 with a different score. Why didn't we follow it? We analyzed our data again, and we found that there was not an independent value between initial FIGO stage and residual disease because all the early stage have no residual disease in frontline surgery, so it's two sides of one coin, and the same was true in our data was CA1 to 5. There's no independent value of ascites and CA1 to 5 because all patients with ascites, a lot of ascites have a high CA1 to 5, so that's the reason why we stick to our score, and that brought different populations to different trials, and Rob, this is a new slide, so please take care, pay attention. This is new. So the desktop trial, all patients had a positive AGO score. In the Chinese trial, 41% of patients had a positive AGO score and 59% had a positive Tian score but not a positive AGO score. The blue percentage of patients with a positive AGO score were not covered in the Tian score. And then they put in another proportion of patients where the PI could overrule the score by saying, no, I believe this is resectable. And it was, and then the patients were, I think. What you see here is that we have the most, the strictest score of all that. The GOG has not a formal score. The recommendations of the protocol were, compared to that, a little bit loose. So the design of all the trials were pretty much the same. Patients were randomized between surgery and no surgery. In our study, the patients had to have a positive score and we had only specialized centers. Again, the score was confirmed within an interventional trial with a complete resection rate of 74%. And we found also very low mortality rate, giving us the certainty that the score also selects good performance state patients which could tolerable surgery. And it confirmed our selection of centers who are capable of doing this type of surgery. So that was a primary outcome. It was a positive first endpoint. So overall survival was improved. Median overall survival by almost eight months with a hazard ratio of 0.75. And that was, coming to the discussion you had before, that was an analytical analysis. And it was pre-specified as a primary endpoint. When we look in subgroups, we found that patients with a complete resection had the highest benefit and that was the 74.2% of patients. And the benefit compared to no surgery was 16 months with a hazard ratio of 0.47. However, you have to be aware that only three out of four patients achieved this complete resection. And one out of four patients ended up surgery with a residual tumor. And these patients have only the toxicity and no benefit at all. And that you have to explain the patients before you discuss surgery. So then comes Rob's study. So why GOG was not a positive trial. There are many reasons which you could think of. We don't know. In fact, we still are planning to put our data together to maybe find more reasons. But I will discuss some of the potential reasons. The first thing is that we had a different study design. However, I don't think that this design has so much impact on the outcome. Maybe center selection has some impact because we only selected surgical centers, specialized surgical centers, as did the Chinese group with Ron Yusang. The center selection in the U.S. was mainly based on the BEV study. And all the large surgical centers were not contributing a lot to the trial. The selection criteria for me is the most important thing. We have different selection criteria and we have the strictest on that. The complete cross resection rate differs a bit but not too much. So this alone would not explain the difference. Surgeon therapies after surgery are different. And the most important is BEVASISIMA with 84% in the GOG and only 23 in ours. So does that explain the difference? And I was always fascinated. My opponent has a huge capacity of working with PowerPoint and you always have these flying curves and all that. So I was completely, completely intrigued by that. And so I learned it. And this is GOG. These are GOG subgroups. Left on the left side you see the subgroup without BEVASISIMA which is a very small subgroup. In this trial only 16% of the patients. And in this subgroup there was a detrimental effect of surgery. And then you have a larger subgroup and this is a subgroup with BEVASISIMA plus minus surgery. And what you can see is that there is only one outlier and this is the patients not receiving BEVASISIMA but being operated. I don't know why. That contradicts all data we have about the additive effect between surgery and drug therapy usually. This is a new slide as well. When you are trying to solve riddles in real world you always can say CUI BONO who has benefit or just say follow the money. In medicine my thesis is follow the evidence. And look at the subgroups we have. Especially for Rob I did this analysis. This analysis is only those patients having not received BEVASISIMA in the desktop trial on the right hand versus this subgroup in the GOG trial. As you can see a much larger proportion of patients being in the desktop trial because only 23% received BEVASISIMA and we have just the opposite result than in the GOG study. In the non-BEV cohort again surgery was significantly better than non-surgery. So this is the truth. And the truth about surgery in recurrent ovarian cancer is follow the evidence. Test scope 3 based on prior work like phase 1 and 2 studies and performed as planned including a validated patient selection process with mature data showed a significant and relevant overall survival benefit. Eligible patients should be treated in competent centers and must be informed about their chance of prolonging survival. But these patients have to be strictly selected. That's my pro. And unfortunately you used so much time with your session before that we must skip the con part. Thank you. Andreas, that was a great performance on a foreign soil against a hometown hero. Yeah. You even came up with speaking to surgeons, right? Yeah. With new RopaDope evidence that Dr. Coleman will have to respond to. Dr. Coleman, what say you? Yeah. So for the five seconds that I have left to speak I just say that that's what happens when you do subgroup analysis of small sample sizes. So here's my disclosures again. So yeah, I'm going to try to provide you some context. And I think actually what we the session we had just before this with the PARP is actually apropos to this discussion. So again we're not debating whether or not surgery is better than chemotherapy. What we're debating is whether or not surgery in the context of chemotherapy treatment or adjuvant therapy does surgery add anything. And when this is focused in a group of patients that are considered platinum sensitive and we did our best among the trials as you'll see to actually normalize for a number of confounding variables. But this is not a new argument. We had this debate in the GOG in 1992 and now it is 2022. So this is a 30-year-old question that we're still talking about. And if you look at the hundreds and hundreds of retrospective studies, cohort studies, and IPTW type trials trying to do statistical modeling, basically this is the conclusion we can come to. Patients that have short platinum free interval or chemo-resistant disease do not benefit from surgery. The disease just grows back. Patients that have diffuse disease are not good candidates for surgery because you can't completely decide to reduce them. Patients that have suboptimal cytoreduction do not seem to benefit. I'll show you why that's the case and you saw some of that already. And like I said, fancy statistical manipulation is just not sensitive enough to make a difference. Now, because of that, we felt that there was an international call to try to come up with a way to assess these questions and there's five trials that you can see that were started. Two of them closed for futility and the three that we'll talk about today went on. So people always forget about GOG213. Andreas mentioned that it was, you know, it was a two, kind of a two-arm sequential study. One of the arms in that trial was actually a true bifactorial design, okay? You saw when he was using this, like, he doesn't understand what's going on with that group that didn't get BAP. That's because that part of the trial was done first and it was done in 107 patients where there was a bifactorial design, meaning that they had to first be randomized to surgery versus no surgery and then they had to be secondarily randomized to Bevacizumab versus no Bevacizumab. That part of the trial actually finished relatively rapidly with 674 patients and then after that was done, we closed the question of the Bev, no Bev and we opened it up to the investigators to make that choice. That's why 85% of the patients or 84% of the patients actually got Bev after we closed that trial, after we closed that arm of the trial. So it's a very unbalanced, sequentially biased population. That's why those subgroups are very difficult to interpret. Now you saw, and I will definitely, I will stand here all day long and give credit to the AGO group for doing the desktop series. It was absolutely, I use it as an example of how clinical questions should be attacked. Looking at the retrospective data, developing a prospective score and validating it and implementing it into a phase three trial. My congratulations to R2, Andres, is the right way to do it. The problem is that in the real world, docs do what they can do based on their own surgical skill and I think that one of the issues we always debate about is that it's very hard to randomize surgical trials because there's such a variability in surgical skill. So GOG 213 was actually set up to assess that question by using the physician's own abilities to select those patients in whom they thought that they could have a complete grocery section. That was the goal. Also the goal for the scoring systems, both done by AGO and being done by the SOC 1, the Chinese. And Andres told you a little bit about how this model came to be and there are all kinds of issues of which I'll address as to that particular trial in which we're still waiting for final maturity. Now you saw this slide that Andres stole from me, but that's okay, I still love you, that basically puts these in context and I want to highlight this because there are a lot of similarities between the outcomes of these trials, predominantly in the surgical arm. You can see in red the difference in the selection criteria, but if you look at the complete grocery section rates, they're very similar. Actually there's no statistical difference, I didn't put that on here, but there's no statistical difference. We do see there are some differences in the patients who actually were allowed to go on to subsequent surgery after randomization. And it's most notable in the SOC 1 trial where more than a third of the patients went on to tertiary or quaternary surgery after randomization. And then he mentioned already that the post-adjuvant therapy was very different. So the primary endpoint for all three of these trials is overall survival. We felt that was the most important endpoint at the time. This is where the relationship to what we're seeing in PARP is also coming back to haunt us because we can't control for post-progression therapy. And that's going to be really critical in understanding these results. Now as he's shown you, I've done this sliding the curves over because you can see that SOC 1 and desktop 3 are very similar. And if you take desktop and 2.13, you see they're not as similar. But if you take out all the noise of the control group and you look at the surgical groups, don't look at the medians. Look at the curves for the entire duration of exposure during follow-up. So you can see that it's not just the medians that are similar. It's actually all of the curve are similar throughout the entire observation period. In the progression-free survival, which most notably should be the biggest impact from surgery because if you think about it, you take a patient to surgery, you cut out their measurable lesion, and you follow them to progression. That's kind of the standard thing. The most important variable for the intervention, if it's surgery versus no surgery, is progression. What happens for the next two years afterwards is all noisy. So let's look at PFS. You can see the curves that are listed here, SOC 1 and desktop 3, pretty much the same. Same population, same outcome of surgery, same kind of treatment afterwards. But if you look this against desktop 3 and SOC 213, what do you see? Well, what you see is that the surgical arms for the two trials, desktop 3 and SOC 1, are very similar. But if you look at the control arms for desktop 3 and SOC 1 are very similar. But if you look at the control arm for SOC 213, you can see there's a very big difference. So now if we look at the surgical population, you can see where we come to consensus. For the patients that you choose to operate on and you do not get a complete gross resection, getting back to selectivity, you can see that all three trials come to the same conclusion. There seems to be a very big difference. And actually, if you look at this in the context of their control arms, there's actually a detriment. So there's a harm in being wrong, which is why Andreas has focused so much on that selection criteria, because if you're wrong, it has a really big detrimental effect. So the bars going down are the differences between the non-complete gross resection patients for who do not get a complete gross resection and go to surgery versus chemotherapy. If you can see that when they get a complete gross resection, there's a big benefit for SOC 1 and for desktop 3. Now in desktop, in GOG 213, we didn't really see any difference in either. But it certainly was worse if we took those patients to surgery and did not do a complete gross resection. So again, selectivity, I would agree with you, is very much the case. So what have we learned? Well, first of all, surgical trials are incredibly difficult. Andreas and I have been doing this for 20 years. This concept of surgery was first pitched to the GOG in 1999. It took us a very long time to do a surgical trial. No adequate control for surgical expertise. We're already talking about this with neoadjuvant chemotherapy, right? So that's why we're doing the trust trial, because we don't believe that the doctors can operate. You know, the selection criteria is that we've, even with the scoring algorithms, are just too blunt because we're not getting consistent results. And the primary endpoint for PFS is contaminated, obviously, by the removal of disease. But again, surgery is not on trial here. This is one of the most important factors that we continue to look at over and over again. If you think about it, if you design a trial that has 90% power to detect a difference in overall survival, if all the patients die on the day they progress, there's a 90% chance that that statistical result will show a benefit in overall survival. But as time goes on, in between progression and death, the power for that original trial is impacted negatively by the duration of time. So where do we find our surgical trials? Well, they're not even on the graph, because they're more than two years from the time of progression to the time of death. So we have lost our statistical power to even assess the PFS benefits that we're seeing across the trial. Now, we did see it in Desktop 3. But interestingly, we did not see it when we actually put all the trials together, which I'll share with you. And this is one of the reasons why. One, we talked about anti-angiogenesis therapy has an impact on progression-free survival. We've seen that. PARP inhibitors have an impact on progression-free survival very consistently. Anti-angiogenesis therapies have an impact on overall survival in GEOGEO213. We published that, right? And SOLO2, if you look at the patient population that did not cross over to a PARP inhibitor, had a significant benefit on overall survival. So these are uninformed by surgery. These are post-progression confounders. And that's really kind of what I want to talk about, because we now also have evidence, if you want to call it that, that PARP inhibitors could have a detrimental effect on overall survival. And this is not to describe that these are the reasons that we're seeing the differences. I want to drill home the concept that post-progression survivorship can confound how we interpret trials that are based on very proximal interventions, like surgery, like Bevacizumab, like PARP inhibitors. And this is probably one of the biggest issues that we have never solved, is what is the interaction between surgery and adjuvant therapy? I can tell you that in the bifactorial part of GEOGEO213, there is a relationship between benefit from surgery and benefit from Bev. And who knows what that impact might be for PARP, particularly in patients who have BRCA mutation. So this is one attempt at this. As Andreas mentioned, we're going to put our stuff together, and we're going to do a meta-analysis. But I want to give you a preview. So this is work my good friend Donna Brennan, who put these trials together. And I wanted to show two things on this. One, if you look at the bottom where it says retrospective studies, you can see the overwhelming benefit from the retrospective studies on surgery. But if you look at the three randomized trials where that arrow is pointing, you can see what the conglomerate of the overall survival hazard ratio is, 0.93, with a hazard ratio that breaches one. So our takeaways. Highly selected patients may benefit from surgery and adjuvant therapy, but it's going to be driven by tumor biology, and that's where we need to kind of go. The interaction of these two interventions are definitely ones that need to be sorted out, and that's going to be our future intent for surgery and adjuvant therapy. And unsuccessful clearance of disease ain't going to help your patient, and there's a big downside for being wrong. So I would agree with you that selection is important. And I hope we don't do this. Just bury our head in the sand and not do this anymore. Thank you. I think for the sake of time, we'll allow you to ask questions up on the terrace where hopefully both speakers will be a little bit later. So that, I want to thank all the speakers today and all the course planners and moderators. It was a great session. So this concludes day two of IGCS 2022, and there is a reception, as I said, on the fourth floor pavilion, out on the patio, I believe, unless it's gotten too cold. All right.

clinical trials

first-line setting

ovarian cancer

neoadjuvant chemotherapy

interval debulking surgery

bevacizumab

PARP inhibitors

molecular subgroups

progression-free survival

overall survival

platinum-resistant disease

immunotherapy

recurrent ovarian cancer

hyperthermic intraperitoneal chemotherapy

secondary cytoreduction