Hello, everyone, and welcome to the master session dedicated to ovarian cancer. I'm Dr. Douglas Levine from NYU Langone Health in New York City, and I'll be moderating this session with Drs. Robert Coleman and Nicoleta Colombo. These master sessions are a new type of programming for the IGCS meeting, and the goal is to hear from leading experts on these topics and cover the latest research on ovarian cancer diagnoses and treatment in one session. We will have a variety of content formats, including didactics, interviews, debates, discussions, and case presentations. The first talk is from Dr. Martin Cobell from the University of Calgary on how to subclassify ovarian carcinoma into histiotypes and molecular subtypes. Dr. Cobell, thank you for being with us today. Please share your screen and begin your talk. Thank you very much for the kind introduction. Am I – oh, yeah. Good day. Am I sharing the right screen? Yes, that looks good. Good day. If you ask gynecological pathologists what was the biggest change over the last decade, most would say the recognition and implementation that ovarian carcinoma histotypes are different diseases. One similar study published in 2016 showed that a simple four-marker immunohistochemical IHC panel can aid in the distinction of the main histotype listed in descending order of frequency. In other words, histotype diagnosis is now objectively supported by four IHC markers, WT1, P53, Napsin A, and progesterone receptor. This would not have been possible without P53. In the same year, there was another seminal paper that showed that optimized P53 can accurately predict the presence of P53 mutations. This study defined three abnormal patterns that correlated with specific types of P53 mutation, including the uncommon cytoplasmic pattern that occurs in 2% of high-grade serious carcinomas. Abnormal mutation type patterns predict the presence of P53 mutation with 100% specificity in this study. In other words, if you see one of those patterns, there must be a mutation present. This is to be contrasted with no mutation detectable in the control cases. However, a few high-grade serious carcinomas, those with splicing mutations or the C-terminal truncating mutation that result in a truncated mRNA not undergoing non-sense-mediated decay, will express a truncated but non-functional P53 protein showing up as a normal wild-type pattern, which limits the sensitivity. Despite that limitation, P53-IHC is now an important adjunct to distinguish low-grade from high-grade serious carcinoma in small pre-treatment biopsies, as shown in these examples. The morphologic features are micropipillary and low-grade, but the immunistic chemistry shows mutant-type P53-IHC, which is incompatible with the diagnosis of low-grade serious carcinoma. So the correct diagnosis is high-grade serious carcinoma. Historically, the distinction between ovarian endometrioid carcinoma grade 3 and high-grade serious carcinoma has been very challenging. And this was reemphasized in the 2018 studies, where the inter-observer agreement regarding histotype in this particular difficult setting was only 69%. The two examples where the observers agreed, and the two examples where they disagreed. When they were asked to reassess their histotype diagnosis with the knowledge of the WT1-P53 results, they either confirmed their diagnosis or they changed it to, in this case here, in these examples, high-grade serious. This increased the diagnostic precision to 96%. Hence a simple combination of WT1 and P53 is now considered the minimal confirmatory panel for high-grade serious carcinoma, with a very few exceptions, as illustrated down here. This can be applied to large cohorts. Just two days ago, the revised fifth edition of the WHO classification of tumors of the female genital tract has been published. It continues to list the main five histotypes and stresses the importance of integrating ancillary immunosymmetry as more objective confirmation for histotype diagnosis. The next level of research is now to determine molecular lab subtypes within the hierarchical classification of histotypes. The numbers in these groups will be getting smaller, and then collaborative research is required. Let me show some examples. For high-grade serious carcinoma total in the TCGA proposed four molecular subtypes by mRNA expression profiling. This recent work here now created a consensus molecular subtypes based on a robust nanostring platform. To me, this is an example of the highest scientific standards and how the ovarian cancer research community can reach consensus with avoiding the chaos we see in other fields. However, a similar stratification can be achieved by a simple IHC stain, and this is a quantification of tumor infiltrating lymphocytes by CD8. While there's biologic information in the molecular subtype, another approach is shown in this study where 101 gene signature also assessed on the nanostring platform was to use to optimize against outcome. And a more continuous stratification in quintiles is shown here with a remarkable difference in five-year survival. While mRNA expression profiling assesses mostly the microenvironment, it's interesting to note that tumor intrinsic factors in this study from David Botell's group showed the combination of HRD with RB1 was enriched in long-term survivors, and even in the subset of exceptional responses which are characterized by long progression-free survival, this combination predicted a higher chance of survival. Using these tumor intrinsic features, an elegant study by James Branton showed the bewildering complexity of copy number signatures shown represented here by a stack bar plots for each patient. High-grade serious carcinomas are high-entropy disease based on a copy number level. So Abshaw's group took another approach and divided high-grade serious carcinoma by the main mechanism into chromosomal instability with HRD high and foldback inversion FBI high. The HRD-BRCA mutated cases are now considered for PARP, which is great, but we also need to turn our attention to the FBI high group, which is lead alteration, which is high-level cyclin E1 amplification, and this is associated with a dismal long-term survival. The minimal standard of current testing is somatic and germline BRCA testing. I just want to point out, if modern histotyping is applied, BRCA alterations are confined to high-grade serious carcinoma. One example is shown here on the right. I was asked to confirm clear cell carcinoma type in this study, and by applying the four marker IG panel on tissue microarray section, I assessed that some tumors are more likely high-grade serious carcinoma, not clear cell, which was consistent with later genomic analyzers, which showed that the BRCA mutation in these tumors were actually in the misclassified or reclassified high-grade serious carcinoma. This recent study on low-grade serious carcinoma, I believe, shows that there are at least five molecular subtypes in low-grade serious carcinoma emerging, which are associated with higher risk of death. The first one is alteration in CDKN2A or P16, PR loss combined with high fraction of the genome altered, a group with MAP kinase mutation, a group with a newly discovered ubiquitinase mutation, and a group with no specific molecular alteration. For endometrioid carcinomas, there are the same molecular subtypes in ovarian endometrioid carcinoma as have been shown for their endometrial counterparts. In this study, molecular subtype shows better stratification than grade, and that was most apparent in the low-stage setting. The irony is that this was the first study to validate grade as an independent prognostic factor within ovarian endometrioid carcinoma, and at the same time, bringing grade to its deserved resting place within ovarian carcinoma by showing that there's no use in the setting of low-stage while molecular subtype has. The question is now how to tie PR, which is strong prognostic factor, and a possible predictive marker into this molecular subtype. If modern histotyping is applied, MMR deficiency is confined to endometrioid carcinoma. In conclusion, histotype classification is now based on an integration of objective IHC markers, and histotype-specific genetic and predictive testing should be performed. So that means high-grade serous carcinoma should be tested for BRCA1 and endometrioid carcinoma for mismatch repair deficiency. This has also important implication for the inclusion for histotype-specific clinical trials, and emerging molecular subtypes require multiple technologies. Thank you. Thank you, Dr. Cobell. That was a wonderful introduction to this session. We will now move on to the next segment regarding ovarian cancer surgery, and Dr. Sven Manner will begin this session with an update on the TRUST trial. Please go ahead and begin your presentation. Thanks, Doc, for the introduction, and thanks, everybody, from the program committee to putting up TRUST and an update on the TRUST study on this year's schedule and program. And let's start with my disclosures, which are listed here. The TRUST study, I think, is a study that most of you might be familiar with. It's another approach to try to answer the question, what is the best for our patients with advanced ovarian cancer, upfront surgery or interval-dividing surgery? The first third of today's session will actually circle around this debate, and after my update on the TRUST study, there's going to be a pro and con session on upfront or interval-dividing surgery, so I will just try to lay the stage for the discussions coming after me. The TRUST study was a very close and intensive international collaboration of different study groups, different countries, and also single centers that we started after the two previous phase three studies that addressed the question did not yield so many results or so conclusive results that we could definitely address and answer the question. So the design of the TRUST study was therefore pretty similar to the two previous studies, so patients with advanced ovarian cancer could be included into this study. One important aspect regarding patient selection in this study was that the patients had to have suspected advanced ovarian cancer, and the experienced ovarian cancer surgeon had to estimate that this patient could potentially undergo complete resection. So we excluded patients where we initially could be sure that those patients were not candidates for complete surgical resection. The reason might be either the pattern of tumor growth or the overall condition of the patient that made extensive ovarian cancer surgery not feasible. So all those patients that could potentially be resected were then randomized and underwent either upfront debugging surgery and a subsequent systemic treatment according to current standards with addition of bevacizumab, parvonibirus, whatever is the standard in the different countries. We did not try to have a medical or a conservative treatment study comparing different chemotherapy regimens, but rather the different surgical approaches in this study. And in the second arm, patients underwent interval debugging surgery, and an important difference to the previous studies is that it was mandatory to first have a tissue biopsy on these patients. So cytologically proven ovarian cancer was not allowed, but we had to have a history, and this is also an important aspect of the translational research projects that I'm going to come to later in this presentation. So that's the pretty straightforward design of the study, and I said that we had quite a low bar for patients to enter the study. Potentially completely resectable ovarian cancer was the only criteria for a patient to get in. However, for the participating centers, the bar was quite high regarding the number of procedures performed in the different institutions, the amount of complete resections in ovarian cancer. So we wanted to select on the one hand as many patients as possible, and most representative patient population we could get. And on the other hand, have on the high volume, high quality centers, and did actually quite an extensive quality assurance program. And this was the hard part with T and also with V in the spelling for the trust study, for the centers to participate, and also for the quality control, quality assurance program during the study. So with this approach, we selected a few handfuls of centers all over Europe, and also one in the UK, and one in the US, which did not fit on this map. So that's why it's located here in the middle of the ocean. But the representative from this center is also actually in today's session, it's Dennis Chee. So hello to Dennis, and thanks for participating here, and sorry for putting you in the water here. And with the joint effort of these centers, we were able to recruit and randomize almost 800 patients over a little bit over a year. So quite successful patient recruitment in this study, and recruitment finished in summer of last year. And currently, we are waiting for the results regarding the primary endpoint of the study, which is overall survival. I think the only viable endpoint for such a study approach and such a question. And this is expected to be ready by 2024, however, we will see if that's the case. And it depends, of course, on the number of events. And there's, of course, a few secondary endpoints that are listed here, and also many exploratory and translational projects that we addressed together with this surgical study that I would like to highlight over the next few minutes and put your attention to. Because I think a surgical study like this gives the unique opportunity to also collect specimens for translational research and perform different translational research projects in such a setting. And if you look at the two different arms, we collected for this trust translational project tissue as well as blood samples. And in the primary debugging arm, we collected blood prior to the surgery, prior to the start of chemotherapy and at the end of chemotherapy. And tissue, there was of course only one time that we were able to collect tissue during the primary surgery. In the interval debugging arm, however, we had the opportunity for half of those 800 patients to collect tissue at the biopsy after anemization before starting of chemotherapy. And then again, blood samples prior to start of chemotherapy, prior to debugging surgery, and after the end of chemotherapy in this concept. Currently, we are in the phase where while recruitment of the patients is finished and we are waiting for the primary endpoint of overall survival, we are getting together all the tissues from the participating centers. So if we want to wrap up the current status of the study, the most important point at this moment is that tissue collection is still ongoing until the end of this year. We have a couple of hundred patients already registered for the translational project and tissue collected. But I'm sure in the audience, at least one representative of each participating center of the trust study is present at the moment. So to those of you who already sent the samples, thanks a lot. And to those of you who are still planning to send the samples, please hurry up. And if there's any questions regarding the translational project, don't hesitate to contact me personally or our study staff in order to get as many samples for this project as possible. So we can hopefully not only answer the question, what is the right timing for surgery in patients with advanced ovarian cancer, but also answer the question, if we can, for example, predict resectability of ovarian cancer in a certain subpopulation of patients. There's very interesting previous analysis on this, but not in the cohort like we have in a trust study. We are looking for specific tumor aspects regarding the response to chemotherapy, which is also possible in this unique setting of tissue samples collected prior to neoatomic chemotherapy and then at interval dividing surgery. And of course, we are very open for additional projects. So if there is anybody in the audience that has some good ideas what else should be done in this translational research project, you're all very welcome to come up with any ideas and send your proposals. And we are happy to discuss and hopefully include them in our project. So that's so far the update on the trust study. And I would like to end by thanking everybody who participated in this study, patients, of course, but especially also the surgeons. This was not an easy task to finish this study. And the speed of recruitment and the dedication of the participating sites and surgeons cannot be thanked enough. So it was really a great success already to completely recruit that study. And this would not have been possible without the surgeons, the research staff, and the patients. And so I'm looking forward to the results of the study. And I'm looking very much forward to the upcoming debate on the pros and cons on upfront or interval dividing surgery, given the currently available data of phase 3 studies and what else we know about advanced ovarian cancer surgery. Great. Thank you, Dr. Mana. That was another wonderful presentation. Exciting results coming in a couple of years. Please feel free to send in your questions to any of the participants. We will have a discussion in a short while. And we'll get to all the questions for all the presenters at that time. At this point, we'd like to have Dr. Anna Fagotti, who will present a component of the neoadjuvant approach to surgical cytoreduction and newly diagnosed patients. And Dr. Rob Coleman will conduct this interview with her. Thank you, folks. Take it away. Thanks so much, Doug. And congratulations, Dr. Mana, on completing TRUST. I think you set the backdrop for this next series of discussions, looking at the pro-cons for neoadjuvant chemotherapy based on available data. And it basically sets up the whole point of why we're doing that, why you guys did that trial. So I'm going to ask Anna to address these five questions in her talk. So first of all, who are the right patients? What algorithm is maybe chosen to triage patients with neoadjuvant? Are there prognostic factors that we need to pay attention to when we look at this cohort of patients? As Dr. Mana just mentioned, are there molecular characteristics that can help not only predict the response to neoadjuvant chemotherapy, but ultimately, whether or not there's in the emergence of chemo resistance? So Anna, I'll ask you to take it away and address these questions. Thank you. Thank you very much. And thank you, everyone, for this invitation. These are my disclosures. So let's start. Results based on this excellent review on randomized data evaluating neoadjuvant chemotherapy versus primary bulking surgery show that all these studies, besides their own limitations, which have been extensively discussed before, show that neoadjuvant chemotherapy is not superior or not inferior to primary debulking surgery. Moreover, all these studies also agree that the rate of major postoperative complications is higher in primary debulking surgery versus neoadjuvant chemotherapy followed by IDS. So in principle, we may state that each patient is a possible candidate for neoadjuvant chemotherapy if with advanced urinary cancer. However, the use of neoadjuvant chemotherapy is largely accepted in frail and elderly patients. And the reason is double. On one side, we want to reduce the tumor burden with the aim to reduce the number of multidimensional resection needed to achieve complete cytoreduction. And on the other hand, we want to reduce morbidity and mortality due to the combination of extensive surgery together with comorbidities related to age. And therefore, we want to make surgery an option even for those patients who are unfit for surgery. And as you can see in the slide on the right side, you see that elderly patients have a higher residual tumor after surgery with lower surgical complexity score, which means that probably we use inadequate surgery for these patients. Therefore, probably modification to the sequence of treatments administered with priority given to neoadjuvant chemotherapy may assure these patients to achieve complete cytoreduction and then complete treatment and definitely longer survival. However, there are some patients who are not frail. They are not elderly. They are fit. They are young. They are supposed to be resectable, but in fact, they are not. And so we, in our department, use laparoscopy before deciding whether or not managing these patients to primary or neoadjuvant chemotherapy. And this is the algorithm that we use, staging laparoscopy. First, we assess for the presence of small bowel carcinomatosis or mesenteral root retraction. Indeed, in our experience, we show it that these are the main limitation to achieve complete resection. Then we assess, of course, diffusion of disease, including presence of peritoneal carcinomatosis, diaphragmatic carcinomatosis. We assess the disease in the upper left quadrant, spleen, lesser sac and stomach, superficial liver metastasis or mental cake, and the need of bowel resection except for anterior rectal resection. We have published some papers where we describe how to properly assess the positivity or negativity of these variables during laparoscopy. And we use laparoscopy also to assess the risk of perioperative complications. So we're using a very simple score, including some simple variables such as ECOC performance status, the site is CAE 125 and tumor load. And as you can see here, each variable has a different score and the score is made based on the odds ratio with respect to the risk of having postoperative complications. So here, at least in our institution, if the score is higher than 10 and the risk of perioperative complication is higher than 30%, including major perioperative complications, of course, we prefer to use NERGVEN chemotherapy. Now, what is the most important prognostic factor in patients receiving NERGVEN chemotherapy? As you can see here, these are 11 studies including more than 2000 patients with advanced ovarian cancer receiving NERGVEN chemotherapy. And you see very clearly that patients receiving optimal cytoreductive surgery have a longer survival with respect to those patients who received suboptimal cytoreductive surgery. And especially those who achieved no gross residual disease had a better survival. And the same is shown on the right side from a large series from French authors. Residual disease at the end of interval debulking surgery is a very important prognostic factor. Now, on this background, what we try to understand is whether patients achieving optimal cytoreductive surgery with a residual tumor between one and 10 millimeters have actually the same or better survival with respect to those patients receiving no gross residual disease at time of interval debulking surgery after NERGVEN chemotherapy. And as you can see here from our series, there is no difference between these two groups at time with respect to progression-free survival. There is actually a difference in terms of overall survival in favor of the NERGVEN group. But of course, here is a limitation given to the bias of selection. Indeed, patients receiving no gross residual disease at time of interval debulking surgery are those patients who responded to NERGVEN chemotherapy. Therefore, we are comparing different groups. And actually, it's not a good comparison. Of course, if we might have some predictive marker to predict the response to NERGVEN chemotherapy, then in that case, it would be nice to manage those patients who are going to respond to NERGVEN chemotherapy and then successful interval debulking surgery. There is a growing body of evidence showing that BRCA carriers show a higher benefit of platinum-based NERGVEN chemotherapy compared with BRCA-wide type in terms of clinical and pathological response, as you can see here. Mutation carriers have a 34% of complete clinical response with respect to 4% of mutation carriers. And the same regards histopathologic response, 46 versus 24%. And also, in our series, on around 456 patients who were tested for BRCA somatic mutations, you can see that among those cases who received NERGVEN chemotherapy, those who did not achieve interval debulking surgery for stable or progressive disease had mainly a wild-type BRCA status with 92.6% versus 7.4%. Now, the last point is, is NERGVEN chemotherapy able to induce any kind of chemo-resistance with respect, of course, to primary debulking surgery? Now, this is our series. We performed a retrospective analysis of 175 patients, stage III and IV ovarian cancer. These cases were matched for three different variables, eye tumor load at time of diagnosis, as assessed by our staging laparoscopy, complete tumor resection, either at PDS or interval debulking surgery, and at least 24 months of follow-up. And as you can see here, we were able to demonstrate a better behavior of recurrent disease in advanced ovarian cancer patients receiving primary debulking surgery with respect to NERGVEN group. In particular, the primary debulking surgery group had lower recurrence rate, had a different type of recurrence, mainly single or discrete lesions with respect to carcinomatosis. They had a longer progression-free survival, and also the treatment options offered at time of recurrence were different. More secondary cytoreductive surgery were offered to the primary debulking surgery group. However, I think that the most important consideration is to find a reason or a molecular explanation to this type of chemo-resistant. And this is a study made by some colleagues in Russia, where they did an assessment of BRCA status in tumor pairs pre and post NERGVEN chemotherapy. And as you can see here, they were able to demonstrate that there was a restoration of BRCA1 heterozygosis after NERGVEN chemotherapy. And this restoration was not related to a reversal mutation in the gene, but however, to the selection of BRCA1-proficient clones that were already present in the tumor itself, but they have been selected due to the pressure from platinum compounds. So having said that, the point is, what is the clinical impact of this finding on our clinical behavior or clinical decision-making? So we know that loss of heterozygosity in BRCA-mutated patients means that they are more susceptible to a very good respondent to platinum compounds. And on the other side, we may argue that these patients who restore their heterozygosity, maybe they may call for alternative treatment options after interval debulking surgery. So in conclusion, I would say that NERGVEN chemotherapy is a valuable option to decrease perioperative morbidity in frail, elderly, advanced and devalued cancer patients, or those with high tumor load, that residual tumor still is the most important prognostic factor at time of interval debulking surgery, that molecular factors predicting response to NERGVEN chemotherapy are emerging, and NERGVEN chemotherapy may select resistant clones to standard chemotherapy. Thank you. Great. Thank you very much for that wonderful presentation. Up next, we have Dr. Dennis Chi presenting the primary debulking surgery component, Dr. Nicoletta Colombo will serve as the interviewer. So folks, please. Hi. Hi, everyone. So Dennis, I have a lot of questions for you. I would like just to start by asking this important question, given that there are published prospective randomized trials that have shown no difference in survival for primary debulking surgery versus NERGVEN, but more morbidity for primary debulking surgery. Why are you still a proponent of primary debulking surgery? Hello, everybody. I guess I'll just load up my slides to answer those questions. I kind of knew this was going to happen. Just hit the icon at the bottom there, Dennis. Yeah, right there. Oh, you passed over it. Yeah. That one. There you go. So thanks to Doug, Nicoletta, and Rob. I thought I was just going to answer questions without slides, but then we decided to have slides and I thought I was just going to go back and forth with Nicoletta, but Nicoletta, please interrupt me if there are questions and you want me to stop, but I was asked to talk about the PDS proponent and I will answer Nicoletta's questions. Here's my disclosures. And I know Sven, my good friend Sven, I was going to put a picture of the ocean behind my background because Sven has me now somewhere in the Atlantic Ocean, but I'm in New Jersey talking from my home. And I was, yes, when Morrison and Kettering did participate in the trust study and he alluded to the fact that this was going to be a debate. And yes, myself and many of the people in the audience and on this panel have had these debates about primary debulking versus neoadjuvant chemotherapy. And for the sake of debates, it's always, we're always told by the organizers, you know, try to be very aggressive. And if you're on the neoadjuvant side, try to say that a lot, a hundred percent of patients you have neoadjuvant chemotherapy and, and Dennis, you know, make it seem like a hundred percent of patients in your institution don't undergo primary debulking. But my buddies in the Axis of Evil, Rob Bristow, Bill Cliby would kind of say, no, probably the truth is somewhere in between that us proponents of primary debulking ourselves don't do primary debulking on a hundred percent of patients. And it's more on the order of 60 to 90%. And we do use neoadjuvant chemotherapy, but I guess, as Nicoletta alluded to in selected cases, we, we, we, we are proponents of primary debulking at Memorial Sloan Kettering, and we would try to debulk it unless there's a good reason not to. So looking at the randomized trials, I, I've shown these slides before. I think everybody knows, so I'm not going to spend a lot of time on them, but the idea of the randomized trials, aren't we supposed to practice medicine based on level one evidence? Well, the problem with the first two randomized trials was if we are to believe that complete gross resection is the goal of primary debulking surgery, then less than one out of five patients got the, the, the assigned treatment. As I've said many, many times, when I give this talk, opening up somebody's abdomen and exposing it to air is not therapeutic. So less than one out of five patients got the, the assigned treatment. And also, as we'll talk about a little bit more, in these cases, maybe it was this case selection of maybe sick or very extensive disease, and the primary debulking's arm, the perioperative deaths were a little bit on the high side, if you ask me. And then in looking at the Japanese study and Anna's study, I have a world of respect and admiration for Anna. And so nothing against her personally, but I do believe that they were also operating on patients with very high tumor load and maybe some periop, some personal medical morbidity. And from what I gathered from the paper, the number of perioperative deaths was kind of on the high side. Maybe this has been modified in the article, which I wasn't able to access last night. I think we can all agree, let's find some areas where there's common ground. We can all agree that whether you're going to do a primary cytoreduction or an interval cytoreduction, complete debulking of all gross residuals of disease is the goal of surgery. So this was a meta-analysis by Chang from South Korea, where he showed that basically the higher the rate of complete cytoreduction, the better the median overall survival will be. And this has been a project that we've been working on at Memorial Sloan Kettering for decades. When Dr. Levine was a fellow through to his time, through to now, that as time goes on, we are working harder and harder to get our complete gross resection rate up. And this isn't even the most recent data, but our data from 2010 to 2013 shows that you can have extremely high survival of close to six years if you are able to completely grossly resect 50% or more of these patients. And I'll show you that we're kind of, we're not kind of, we're in this area now with the complete gross resection, but obviously we don't have that overall survival just yet. So there've been multiple studies from Nicoletta's group, from the Mayo Clinic, from South Korea, from Memorial Sloan Kettering that have shown that if you improve your surgical outcomes, improve your complete cytoreduction rates, you will improve your median overall survival. However, this study out of Sweden was just published this year and showed that when they changed their paradigm to include upper abdominal surgical procedures, they did not improve their median overall survivals. And as a matter of fact, those patients who had extensive surgeries didn't do quite as well as we would have predicted based on prior studies. But this is just one study. And I think the problem is that, as C. Paul Morrow said, the shortest survival is a perioperative death. So again, patient selection is very, very important when you're deciding on whether to do primary debulking or neoadjuvant chemotherapy. Should everybody then therefore undergo universal laparoscopy? I think it's a great tool. And I'd like to say that nobody should have laparoscopy. But as you'll see from our own data, we do believe laparoscopy should be used in these cases. The problems are you can miss disease, especially in the retroperitoneum, if you only use laparoscopy. So CT scan is important in that area. It can overestimate the extent of disease. And you can think you have small bowel carcinomatosis over the whole area. Dennis, can I stop you for a minute? I want to first, you're now moving to the selection criteria. I want just to go back to the question about randomized trials showing no improvement in survival but less morbidity. So if I go to, I mean, this was a scientific explanation, but if you go to a patient site, so which argument would you use for a patient who comes to you and says, Dr. Chi, you are a wonderful surgeon and you are proposing me this aggressive procedure with a lot of possible complications, but my neighbor with a similar diagnosis just got three cycles of chemotherapy and then not even an open surgery, but she underwent a laparoscopy, a robotic surgery with four small holes and no complications. She's doing perfectly fine. So why should I go through a devastating surgical procedure instead? What would you tell this patient? Well, I would tell the patient, I think that's a great point. And I think that the way that you can have no surgical complications, this is an exaggeration. The way to have no surgical complications is not to operate. The way to have less surgical complications is to do a laparoscopic biopsy. The way to have a little bit more complications, but do a little more is to do a laparoscopic hysterectomy and omatectomy. The way to have a little bit more complications is do a laparotomy, et cetera, et cetera. So the point of what we've been doing for the past 20 years is that you can do surgery and if your metric is surgical complications, then less is definitely better and less is definitely more. But the problem is if you do that and you don't do the surgery now or don't do the primary debulking to a complete resection, you're going to pay later. And that's why I didn't have the time to put in the slides. There are no studies that I know of that have done neoadjuvant chemotherapy in advanced ovarian cancer patients and show median survivals over 60 months or so, or even 50 months. Most show median survivals in the 40 month range. So sure, you don't have to have a big incision. You don't have to risk the complications. But in the end, almost every single study that I know of that has looked at primary debulking surgery in institutions, like Sven has included in the TRUST study, have shown median overall survivals of over 50 months for all comers, including ours. And that's why I showed that one meta-analysis by Chang. And basically the long median overall survivals, yes, it's due to a little bit of patient selection, but it's due to the fact that you're doing the surgery first when you still have six cycles of postoperative chemotherapy, rather than using three cycles of chemotherapy to shrink the tumor and make the surgery less morbid and less extensive. Okay. So going to the selection criteria, since we have only a few minutes, I want just to ask you direct questions about the selection criteria. First of all, do you use a frailty score and which one, if you do? The second question is, do you use a pre-habilitation program for patients who are frail before surgery, before primary debulking? Well, just to answer that. Can you say the first question again? Yeah. Do you use a frailty score to assess the patient's status before going into primary debulking surgery and which one do you use? Or you look just at the former status, age, comorbidities and so on, but you put them together in a frailty score or not? So we use, the Mayo Clinic has published a lot of papers on this and we use the paper, we participated in the paper by, where Aleti was the lead author who worked with you. And basically the Mayo Clinic has published now at least three papers on this, where they use to determine whether somebody is a candidate for surgery, they use a combination of age, the extent of disease and essentially the frailty, the ASA, and a big factor in the Mayo Clinic data is to use serum albumin. So you use those criteria from Mayo Clinic, basically. We use the first criteria from Aleti that we participated in way back in 2010. And yes, our postoperative mortality is less than 1%. And I think that plays a major role in why our patients are able to do well. So now this is the patient side. So in terms of tumor resectability, if you have a pleural fusion, do you perform VATs always or when? Yes, we do. So in a patient who has a pleural fusion, we have found that we find solid tumors 60% of the time in the chest. And the patients who have gone on before we had complete gross resection, as the goal, we used to do primary debulking, even if there was disease less than one centimeter in the chest. And we're about to publish a paper, it was just accepted by GYN oncology, of the first 100 patients. And the bottom line in that paper is that if you had gross residual disease in the chest, it didn't matter what you did in the abdomen, those patients' survival was under 40 months or so. If you did, if they had no disease in the chest and you completely gross resected the chest and the abdomen, those patients had median survivals in the 70-month range. Okay. So if it is resectable outside the abdomen, you do, even though it's not in the abdomen. If it's not resectable, but small, I'm talking about supra-diaphragmatic disease, but very small, you still go ahead with abdominal debulking. That's what we used to do. Now based on our recent data, if we cannot clear the chest, we don't operate on the abdomen. Even though the disease is very small in the chest. Correct. Yeah. You don't. That's been a change. That's something that we changed just based on this recent data. Yeah. So basically you believe that stage four disease is a bad prognostic factor, but you can rescue this prognosis by doing debulking surgery when it is feasible, even outside the abdomen. Is that your point? Yes. We published a paper that showed that if you have supra-diaphragmatic lymph nodes and you respect them, we had a median survival over 80 months in stage four. You rescued stage four to stage three, kind of. I mean, that's what you are saying. So the last question, because we are over 20 years after the paradigm shift towards aggressive primary debulking surgery, and still in the majority of clinical trials, even presented yesterday, the percentage of optimal cytoreduction remains 30%. So why? Because we still have serious doubts about the possibility or the prognostic factors, or because there's a lack of expertise and a lack of dedicated centers capable of performing this surgery. Yeah, I think it's multifactorial. Do the surgeons want to operate for six to eight hours? I have back pain. I have neck pain. I have foot pain. Is the institution willing? I've heard this from institutions in Canada, Dennis, what you do is great, but there's no way I can do six to eight hours of theater time and do one case when the hospital wants me to do three endometrial cancer cases. I think it's a multifactorial problem. I guess I have to be honest, and the benefit of Memorial Sloan Kettering is that the infrastructure and the philosophy there is, yes, don't back out and try to do the most that you can surgically. But again, we have pretty good results because of the collaboration with other surgical services and the philosophy that we're going to do what we can to get it all out. Thank you very much. I think we are over our time. Totally great. Thank you. Thank you for the discussion. Great. Thank you both for that wonderful interview and a beautiful interview format. We have a number of questions, so I might ask all of the presenters for this session to come back online and the moderators, interviewers, because we could have a collaborative discussion here. Since we just heard from Dr. Chi, maybe just a few simple questions for Dr. Chi. We have about 10 minutes or so for questions. For the 261 attendees, please feel free to send your questions in through the Q&A box and we'll try to get to as many as we can. Dr. Chi, when you do a VATS procedure, and you can stop sharing your screen possibly. I don't know if you can stop sharing it. Maybe the coordinator can do that. When you do a VATS procedure, do you do it stepwise or at the same time as you're debulking procedure? Can you hear me now? Yes. I personally don't do a VATS procedure. We ask the thoracic surgeons. I want to make that clear. I don't have privileges to do VATS, but yes, we have the thoracic surgeons do it. It can go either way. Sometimes we ask them to do the VATS procedure and we're on standby. If they can clear the chest, then we're on standby and we do the procedure. The problem with that is if it's a no-go for a laparotomy, then you've just reserved seven, eight hours of OR time and then it gets lost. More frequently than not, we'll do a two-step procedure where we'll have the thoracic surgeons do the VATS, drain the fluid if they have to do intrathoracic side reduction, place a chest tube, pleurodese, and then maybe in a week or two, do the debulking if the chest is cleared. Great. One more question for you while we're here. You report very nice and impressive low morbidity and mortality rates. Would you say your patients are different from the general population, this person is asking? Can you describe the socioeconomic status, insurance status, demographics, and frequency of BRCA mutations? Yeah. I just can't help. I hope Rich Barakat is listening. I just can't help but quote Dr. Barakat, yes, it is true. At Memorial Sloan Kettering, we only operate on young supermodels. That's a joke and it's not true. Yes, maybe everybody has insurance. I've heard that before. I don't think we have ... Yes, maybe there's more Jewish patients in New York, but the incidence, as far as I can tell, of BRCA mutations in our patient population is the same as everywhere else. It's about 15%. Yeah, but at the flip side of that, absolutely, we use patient selection. I would go for 10 hours to do a debulking on a 40-year-old with two kids who's healthy. I would give neoadjuvant chemotherapy to an 85-year-old in a wheelchair. Absolutely. Patient selection is key and I would never say anything different. Great. Thank you. Maybe for some of our international colleagues, I know Dr. Chi talked about selection for frailty using the Mayo criteria and albumin levels, et cetera. Dr. Fogotti and Manor and Dr. Colombo, how do you choose patient selection based on frailty and patient tolerability of a large operation? The criteria that Dennis described, I think, are used internationally and something I didn't highlight when I gave the update on the trust study is that we actually had a sub-protocol which was not running internationally, but only in Germany, where we tried to compare all these different assessment tools for frailty and the results of this should be ready by the end of this year, hopefully, and so we can hopefully shed some light on the different frailty scores, scoring systems, et cetera, soon. Maybe one addition to the question of stage four disease in the thorax, and of course there's many patients where you would do a VET prior to planning any debilitating surgery. However, it's also quite feasible to do this in an open fashion at the beginning of your abdominal surgery to just open the diaphragm and see what's going on inside if you don't see any larger tumors and only have a fluid fusion, and this is a way that we handle this frequently. Yeah, we do use the frailty score also, according to the Mayo, and also very recently, that was one of the questions Dennis didn't have the chance to answer. We introduced a pre-habilitation program for frail patients in order to be able to go to DOR in a better shape and hopefully to reduce the complication, so I think also this is very important, of course, to pursue, but if I may, I want to ask you a question about the thoracic tumor. This is one of my ideas, so let's say that you cannot devote the supra-diaphragmatic disease, but would you ever consider to do a radiotherapy, a stereotactic radiotherapy and treat the supra-diaphragmatic disease and perform a debulking through a stereotactic radiotherapy instead of surgery and then complete the debulking intra-abdominally? I know that it looks a little bit crazy, but we are having wonderful results with the localized radiotherapy in case of relapse, and so my idea is sometimes you could consider to do a debulking in a different way, so surgically into the abdomen and then using a different technique in different sites. Is this only my crazy idea? Who was that question for? Everyone. If you can hear me, I think that's a great idea, sure. I mean, however you can get the cancer out, whether it's radiofrequency ablation, as patients always say, laser, cautery, ABC, whatever you can do to get it out, I think it's reasonable. I didn't know the data that Dennis referred to regarding their experience with resecting and not resecting thoracic disease, so our bar is quite low to if we can completely debulk the patient intra-abdominally and have only small residual disease in the thorax to go ahead, given the experience that most of the patients dying from ovarian cancer usually don't die from problems in the thorax, but in the abdomen, and so that was the rationale until now to go ahead with any more or less extensive surgery also in the abdomen. Nevertheless, now, given your analysis, we maybe need to reconsider this. Thank you. A question for Dr. Fugate. In the PRIMA trial, two-thirds of the patients had neoadjuvant, but in the subgroup analysis showed no detriment in the efficacy of norepinephrine for maintenance. Any thoughts on that? Yeah, we may argue two possible explanations, but this is only my opinion. One is that most of the patients actually had a successful interval debulking surgery, so at the end of the story, all patients had no residual tumor, including surgery and adjuvant chemotherapy or neoadjuvant chemotherapy at the end of the entire course. So, Niraparib is able to work on no residual disease at the end. The second possible option is that if we believe that there is an heterozygosity restoration after selection with adjuvant chemotherapy, anyway, Niraparib is able to work even in wild-type patients, and it has been demonstrated. So, maybe this might be another explanation, which is different from the use of other types of PARP inhibitors. Yeah, that's an interesting point. And along the same lines, for BRCA carriers, we've seen such an impressive response to maintenance PARP therapy, and we've seen your data about debulking in BRCA carriers, which probably predated what I would call the routine use of PARP inhibitors now, at least in the maintenance setting for BRCA carriers. So, with that dramatic improval in survival and what many of us would believe a dramatic improval for primary debulking patients, how do you coordinate? I mean, obviously, the BRCA patients will do the best, but could you make the argument maybe they don't need as aggressive debulking? I'm not saying that's my argument. Do they don't need as aggressive debulking? Because they're now going to get PARP inhibitor therapy, which is also going to be a form of pseudo-aggressive debulking, so to speak. Any thoughts from the panel on that? We need to do a prospective randomized trial on this. Well, the evidence that we have in many of these trials, particularly in the trials with SOLO1, we did also subgroup analysis. The evidence is exactly the opposite. The best population, the one which may derive more benefit, is the population of patients with, say, three optimal cycle reduction, no residual tumor, and they're doing absolutely great with PARP inhibitors. Even if you compare with the control, so good prognosis, say, three optimal cycle reduction, there is a huge difference with the use of PARP inhibitors. So, I don't believe that you can make the argument that you don't need debulking surgery because PARP inhibitor can rescue the lack of good surgery. I don't really think this is the case. This is especially true for upfront debulking surgery. Yeah, yeah, upfront. I think it's important to highlight that. I mean, I think we cannot highlight enough that the majority of indirect evidence, even from other studies that did not address the question of optimal timing of surgery, shows that the patients do better with successful upfront debulking surgery as compared to interval debulking. I anticipated this question, so I actually had the data from SOLO1. In the patients who had elaporib, the three-year progression-free survival for the primary debulking arm was 69%, interval debulking 47%. In patients who had surgery, the three-year progression-free survival for complete gross resection was 65%, and for those who had gross residual disease, it was only 48%. So, I don't think elaporib will take over for doing surgery. Yeah, I presented this data at ASGO, and now we are updating, and we are presenting new information about this with longer follow-up. Yes, actually, we had the same results also at time of primary recurrence of iron cancer platinum-sensitive recurrence. So, that's true. The point is that probably not all pipe inhibitors are the same, and we are discussing whether they may add something to those people who cannot achieve PDS. So, this is another point. Yeah. I mean, again, I mean, the one thing we got to always remember is that the patients that are selected for neoadjuvant chemotherapy versus those that are primary debulking surgery are already selected for prognostic factors that aren't necessarily overcome by a pipe inhibitor. So, what you would need to do is essentially a bifactorial design, right? Take surgical patients, good surgical patients, randomize them like the trust patients, randomize them to surgery versus neoadjuvant, and then randomize them again to PARP versus no PARP. That's the only way you can really ask this question, because we are all talking about biological effects that are not controlled for by the patient selection of neoadjuvant versus primary debulking. Despite that, we don't think we are going to do the trust trial again, I think. No, I had no idea. I have no idea, too. I was just suggesting that. That's enough. That's enough. By the way, congratulations on completing that, Carol. That is phenomenal. Surgical trials are so hard to do. Well, this has been a really great discussion. We will move on shortly. I think the panel and probably most of the audience would agree that when newly diagnosed ovarian cancer patients can get to a center of excellence to have complete resection with proper selection criteria, as has been very eloquently discussed, that certainly will result in the best outcomes for our patients. It's what we're all trying to achieve. I'd like to turn this over to Dr. Coleman, who will take care of moderating the next session. Thanks so much to everybody. That was a really fantastic discussion. I know it could have gone on for much longer. We really appreciate that. What we're going to do now is another kind of strategy to this debate. We're going to talk a little bit about some clinical vignettes that will incorporate some of the components of surgical and, importantly, the use of adjuvant therapy and how we approach that. I'm going to present. We have three speakers, three wonderful speakers. I'm going to introduce each one of them with a clinical vignette and then allow them to go ahead and present how they would approach these specific patients. The first one is going to be from Dr. O'Malley. The case here, as you can see, is a 58-year-old woman who presents with a six-month history of accelerating early satiety, bloating, weight loss, and most recently, abdominal distension. Her past medical history is significant for hypothyroidism. She has obvious ascites. You can see in the CT scan. She has this pelvic mass. She has this floating, kind of bloatable mass. It's in the mid-abdomen, which you can see there is likely the omentum. She's got a CON25 of over 4,500. She did undergo a core biopsy, and it came back as high-grade serious adenocarcinoma suspect GYN origin. Because the tissue was available, she did have an assessment for BRCA and HRD. She actually came back as a wild type for BRCA mutation, but was found to have homologous recombination deficiency. I'm going to turn this over to Dale and stop sharing here and let you go ahead and take the ball. We can't, I can't hear him. Mute. Can you hear me? Yeah, there we go. Now. Oh, so sorry. Look at that. Sometimes I shut off my Speaker there, so sorry No, thank you very much for letting me that time to discuss this and I'm honored to be presenting for my Attic office in Columbus, Ohio And not enjoying the the world's third largest cancer hospital Today, so as we look at this and these are my disclosures I Really want to spend a little bit of time on BRCA, even though this appears to be a neoadjuvant patient I know I know Rob and Rob doesn't operate a whole lot. So I'm assuming this is a neoadjuvant patient We have a BRCA BRCA wild type in HRD but I think we really need to look at some BRCA positive and use this as as we look at our Patients talk a little bit about genetic instability, which we've referred here is hematoma recombination Deficiency some people refer to it as HRD positive and also look at other other homologous recombination mutations, so Let's start with Bev though And really the question I was asked is Bev versus PARP and I added versus both I think that's really the discussion and we have two New England Journal Medicine articles that really described in Shaped the way we use Bev in the upfront setting what's better than two New England Journal Medicine articles for New England Journal Medicine Articles and now we have these upfront trials first by solo then followed by Prima Paola and bilia As we look at these patients, we have to really kind of think about and put into context put into context that Sorry about that put into context where we are in our overall population. This is the original Presentation which included the censoring for ca-125. These numbers are slightly different. You just look at Rhesus criteria, but we see six month improvement in the placebo versus the bed throughout Arm, and now it's important that when we look at these median progression free survivals We remember this is during the time of chemotherapy and as we'll discuss in solo one There's almost seven months From the first chemo until randomization. So this is as we put these into context It is a slightly different population of patients now As we go through this we know now in the final analysis There's no difference in the overall survival in these groups, but an unplanned sub analysis We did identify a high-risk group of patients different than icon 7 but in stage 4 patients, which did show Improvement approximately 10 months in overall survival and again the bed throughout arm Now as we look at icon 7, we know the well-described We know the well-described high-risk group that include neoadjuvant suboptimal to Balkan as well as stage 4 disease and once again, we see a Survival advantage in this group and this shape has shaped many of us in our utilization of Bev in the upfront setting So as a brief summary, we know that Bev is indicated throughout the world in the upfront first line setting but we have survival advantage in patients in the US population in stage 4 disease and Throughout the world and icon 7 in this higher risk population What's the biggest challenge we've never been able to overcome with Bev which is no biomarker, which we've all been asking for Now I included here the three all-comer trials Prima Paula and Vilia and as we know Crosstalk comparison is not valid and what am I going to do throughout this entire presentation, but use cross trial comparison Rob I love you, but the lab prep does not seem like it's going forward So we're really going to concentrate our discussion on Prima and Paula and for the yo's you don't know. Dr. Coleman was the PI of Vilia We look at the intent to treat population Unbelievably consistent both of these very different populations and we have time to discuss that But we you know with a hazard ratio of 0.6 with improvement of about six months But as we look at the subpopulation this HR Deficient group once again about 0.4 to 0.3 of a hazard ratio. But now we're looking at a Year plus of improvement, but really that's driven by that's driven by this BRCA Positive patients. How about when we just look at these HR Deficient patients and we exclude the BRCA mutations We still see about a year improvement with hazard ratio of 0.4 to 0.5 So again, we take this subpopulation. We look at the sub analysis. These are exploratory sub analysis We can't make too many Jumps here, but the FDA did make a significant jump here with their approval of the combination let's remember solo one really changed how we look at these patient is probably the most impactful trial of our lifetimes of and looking at marked improvement 70% improvement in the progression-free survival and really more than doubling our progression-free at three months We're not as you all know not able to even compare our median progression-free survivals as the treatment arm has not been reached As we look at Prima, I'm going to concentrate obviously on the HR Deficient group on the left. We see the group included with the BRCA mutations again that has your ratio 0.6 I alluded to clear difference, but those differences in impact really persist as we look at the BRCA wild type HR deficient patients Now when we look at Paula, this is the intent to treat population again a modest improvement Median progression-free survival of a six months with a hazard ratio about 0.6. Now, it's very important I think we often forget when we're discussing Paula. This is the only trial that had an active Comparator or historical control which is bad throughout or bad maintenance in this setting since it was a maintenance trial Let's look a little bit more closely at this HR D positive group in Paula That's very important to note that the the the genomic instability test used here was Marion My choice with a cutoff of 42 The challenge with the HR D Marion my choice is we don't get the mutational analysis and I think that is so important as we're really identifying those patients who we Absolutely should be using PARP or some PARP maintenance therapy You see on the right side of the screen about about one half of patients were HRD Deficient, excuse me deficient HRD deficient and of that about two-thirds of her bracken about one-third or 20% entire population was wild type BRCA with HRD Now as we look at this group HRD Deficient patients and the non BRCA patients we see again about a 12 month improvement. Here's the curves This is looking just at that population Again, the active comparator arm here is Bev We're seeing marked difference almost doubling at two years of progression-free survival Very interesting analysis by dr. Goat more and rake record Looking at a propensity test presented this last year's SGO Trying to compare Paula and solo one Looking at that group of patients, but how can we deem more and get more information from this now? There's a lot of challenges propensity waiting trial, but it's a very acceptable Statistical analysis in a way to really compare subsets admittedly admittedly. These are BRCA positive patients But when we look at the Bev versus placebo arm in these two patients, so on the top we have the Bev Clearly while they're receiving the Bev versus placebo we see difference between those arms As we stop the Bev we see those arms coming back Closer together like we've seen in all of our trials But you see a numerical difference at 24 months of improvement in overall hazard ratio which was statistically significant in this propensity analysis looking at 0.65 or a third improvement Now as we look at the elaporate versus Bev, okay, so this is a lap rev again BRCA patients Okay, we see that clearly the PARP inhibitor seems to have Improvement over Bev in these patients and as we put all of this together We see a numerical difference though not statistically difference of the combination in Over PARP am I willing to say that these curves are clearly different obviously not but it's an interesting analysis that you did have a Numerical difference of nearly 10% at two years and again those curves coming back together over time So I want to spend a few extra minutes really looking at what's driving these results in the HR deficient patients who are non BRCA and How much is the Bev really adding here? clearly the PARP Combination is beneficial now I'm using aerial three as a potential example of the importance of having next-generation sequencing of The somatic mutations these patients. I know this is a recurrent setting but in a pre-planned analysis and actually these were stratified for the 32 HR deficient or homologous recombination repair genes We know that the rad 51 C and D and probably 2 has clearly been associated with some clinical sensitivity to PARP And I have the honor of presenting this unplanned sub analysis at SGO this year and what we found is in these patients non BRCA HRR gene mutations about 8% of the patients Harbor one of these mutations as we look more closely at the progression free survivals of this small subgroup We see a doubling of the median Progression free survival the hazard ratio 0.2 in small numbers, but you really when we look at the swimmers plot We're really seeing that these patients Really very few in the placebo arm Make it progression free beyond six months and almost all of them are censored prior while the majority of patients in the active treatment arm and recap rib See improvement and we look at the rad 51 C and D gene mutations It's actually almost a two-year or a year and a half. Excuse me improvement in progression free survival while the other non BRCA Non rad 51 C and D is more modest at about three months and again now We're down if we're looking at the rad 51 C and D about 3% of the population But you see that eight of those ten patients had 12 months or longer Progression free survival and about half of them continue to be on treatment at the last censoring so Utilizing that data. What would I do in this patient Robin? You really asked us and challenged us to say, you know put down what you're gonna do on the order set Well, I'm not I can't be that definitive, you know me and I'm clearly gonna recommend a clinical trial we have three clinical trials at randomized prospective trials accruing throughout the world and we Absolutely need to get all of our patients onto these clinical trials to answer these important questions If she wasn't a clinical trial candidate, she recently been diagnosed with breast cancer. We all see those patients Well, I'm gonna start bad particularly with what her disease is looking like on the scan That's quite an impressive scan with the mesenteric involvement. I'm gonna get that somatic and germline testing right away We don't play around with that. We get tissue sent off now at times We're challenged with regards to amount of tissue We can get on a core and we may went wait for interval tumor reductive surgery If that patient if that patient is found to have HRD in this case I probably will add the PARP if she's BRCA positive or rad 51 C or D because I think that's our best information But we need more information. We have to have the mutational analysis. I was reviewing the prostate data They have jumped ahead of us regards to your mutational analysis. That's so embarrassing. We need to get away from these Analyses of HRD and really look at the mutations in that scenario if they were positive If they had or started a hint to toxicity of the Bev, I would drop it I've talked to many of the colleagues on this call this morning around the world and they don't see some of the Hypertension and real dysfunction that we see here in the Midwest with our patient population of Bev But that's really an issue many of our patients have undiagnosed hypertension and can be a challenge as we use Maintenance Bev. So thank you very much for the time. I really appreciate the opportunity Thanks so much Dave great discussion, you know, I was gonna pin you down you can go ahead and stop sharing if you wish so so We'll come to some Q&A at the end, but I had to I have to interject You do you normally wear your lab coat in your attic office? Oh You know what? I'm lucky. I have I'm lucky I have pants on that's all I have to say You know in with the attic with nothing on the walls I had to do a little signage, you know what I mean Appreciate it. All right, I'm gonna share the next case and I'm ask Shannon Weston to address this patient So I hope you guys can see this So this is a 47 year old so younger patient present an outside institution with acute advent she had previously been asymptomatic Her family history is significant. She has a mother with ovarian cancer age 72 She was a she had a kind of a bedside ultrasound which you can see on the right Which shows this complex bilateral masses and her steven 25 was 254. So she underwent an emergent kind of Laparotomy she had bilateral ovarian masses were identified. The right one was ruptured and bleeding at this point she They noted that there were a few nodules in the abdominal cavity so in the process of stabilizing her they removed these nodules which were easy to obtain as I'm told and So she had a small amount of residual Diffuse disease and they did notice an implant on the diaphragm that they they decided not to go after It was about one and a half centimeters the and they closed and then Waited the final pathology. It's final path came back as a high-grade serious. I know carcinoma the tissue was sent and these are the data it was came back as a An HR proficient case as well type for BRCA well type and negative for HR deficiency So an HR proficient BRCA wild-type patient. So with that I'm gonna tee that up and send it over to Shannon to To discuss So while I pull my my slides up, I just want to thank the organizers. Dr. Colombo, dr Levine and especially dr. Coleman for giving me this case Here's some wonderful data from dr O'Malley about all the things we can do all the options we have that are successful for patients with the RCA mutations or homologous or combination deficient tumors, but I'm gonna tackle the tough one Homologous or a combination proficient disease But frankly, it's it is an important case because this is where the majority of our patients are gonna be you saw dr O'Malley share The numbers and these are my disclosures These are you know, these are the patients we're going to see and unfortunately, we just don't have an ideal option for them yet It's not definitive But I am gonna follow the rules and I am gonna be definitive and that will be just another little Difference between me and dr. O'Malley aside from not wearing a white coat in my office So first let's just review some of the key points in the case and what we'll use to To make this decision for the patient in front of us First of all, we have a stage 3b high-grade serious ovarian cancer patient and her tumor is HRD negative So homologous or combination proficient and we do have residual disease and that's important There was at least at the very least a 1.5 centimeter implant on the hepatic flexure on the diaphragm and so the first question really is How are you gonna treat this patient up front? And I you know, I love this We always do these case reports and or these case discussions and and everyone always says, you know, and I'm glad for it We'll put the patient on a clinical trial. And yes, that is absolutely the right answer, but today clinical trials aren't available So what are we gonna do with this patient in front of us? And and you already saw the these data from the frontline trials that really made bevacizumab in combination with chemotherapy The standard of care, but you know, I'm gonna put it in the context of this patient in front of us today so these were the the curves for both the GOG 218 study as well as icon 7 that demonstrated a Difference in progression-free survival, but you know, it it has always been argued. Is this a clinically relevant difference? Is this something that we really want to change practice with and and is this something that we for three months or six months? If you're using ca-125, you know Is is that enough to put a patient through additional drug with additional toxicity with additional cost and additional time? So I think this is one of the reasons why many sites didn't really in you know, utilize this in in upfront patients However when we look at the overall survival and again bringing it back to the patient in front of us now We know there was no difference in overall survival in the all comers populations for GOG 218 and icon 7 Dr. O'Malley just discussed the kind of and I agree with him this very interesting Overall survival benefit for those patients that are really high risk that really that curve that stage 4 patient curve Mimics the stage 3 curve when you add bevacizumab to chemo in that patient population So that is definitely something that I incorporate when I have the patient in front of me now for this patient though This is a stage 3 this patient has a stage 3, but she does have residual disease so how can we use this to make our decision while looking at the overall survival data and icon 7 and that that high-risk subgroup that they Defined which was that residual disease more than 1 centimeter or stage 4 now those patients did see a survival benefit So perhaps high-risk disease is the biomarker we can use. I agree with dr. O'Malley We need a molecular biomarker and I'm so grateful for all of the translational scientists that are working on this But we don't have it yet. So instead we need to use these clinical biomarkers and you know potentially consideration of this high-risk group this residual disease and And stage 4 patients this is how we can make our decision So I've made a decision because I'm being definitive today and we're going to utilize The combination of paclitaxel given every three weeks with carboplatin and bevacizumab and we'll start that bevacizumab on cycle 2 Now luckily this patient does I'm gonna take it to the next level Rob didn't give us this information But I'm gonna say she receives she achieves a complete response at the completion of therapy So now we have another question in front of us. What are we going to give this patient for her maintenance strategy? So let's look we talked about the bevacizumab data. Let's look at the PARP inhibitor data And yes, we are doing a little cross trial comparison and dr. Coleman I truly love you because I am including Velia in this slide but I think it's important just to see that you know, we do have Information about how PARP works in those HR HR proficient tumors and and the answer is It works a little bit It is not as profound as what we see with tumors that have homologous or combination deficiency And it's certainly not as profound as what we see for those tumors or or patients that have BRC mutations, right? So we know that there could be some activity of PARP, but it's not going to be as profound. So you can see across these three trials, we have very consistent data, right? So in the Velia trial, those patients that had homologous or combination-proficient tumors, there was about a 25% risk reduction. Now you have to be careful, and I didn't say this right away, but I'll say it now. These are all exploratory, right? These are all exploratory analyses. None of these were the primary analysis for these trials. So they're hypothesis-generating, but you're making me be definitive, Dr. Coleman, so I have to use something to make my decision. So when we look at Prima, we see, again, about a 32% reduction in the risk of progression in those patients that were treated with norepinephrine versus placebo, and it yielded about a 2.6 or so month difference. And then finally, Paola1, which has already been discussed quite well by Dr. O'Malley, there was really no difference in those patients that had homologous or combination-proficient tumors, and that's because there was an active comparator. These patients weren't on placebo alone. They were receiving bevacizumab, and the bevacizumab arm did very well. Now, the Prima data, it's important to note that this really does reference, the data that we see from Prima really does reference the patient in front of us, right? This patient had high-risk disease with residual tumors, and that is what the patient population was for Prima, so that's exceedingly relevant. So how are we going to make these decisions? You know, there's a number of different things we think about when we're talking about maintenance therapy. Obviously, we want to be thinking about what's the best outcome, but if all outcomes look about the same, we have to use other factors. And so certainly, you consider the indication. Is the drug indicated in this space? We talked at length today about the BRCA mutation status and the homologous or combination-deficient status. Obviously, those things are going to be considered, but the other thing is toxicity. What are the expected toxicities of the maintenance therapy? What are the existing adverse events that the patient is experiencing? Those types of things can help us make that decision. Schedule is important. You know, does the patient want to take an oral medication? Does she not want to come in for infusions? These are questions that'll help us make that decision. And then cost. I think, you know, I don't get into this today, but there's a lot of really wonderful presentations that have happened over the last few meetings. We've seen them at SGO, we've seen them at ASCO, and here at the IGCS, looking at the financial implications of continuing these regimens compared to what is the potential survival benefit. So we've got to consider all those things. And here are just some other things that help me make these kind of decisions, not necessarily relevant for this particular case, but really important to discuss. So certainly, you know, one of the pros for bevacizumab is we've seen over and over again, when we give bevacizumab with chemotherapy, it improves response rate. Now, I know, I know, that maybe the progression of resurvival isn't as, you know, bountiful as we would hope, or we don't see it reflected in the overall survival, but it does get us to that maintenance decision. So I like that idea for these patients that are high risk. When patients have a lot of ascites or pleural effusions, it's really great to utilize bevacizumab to help dry that up quicker. And if we really stick to bevacizumab maintenance, we do get a biomarker, right? Especially in that homologous recombination-perficient tumor. You know, we get that platinum sensitivity. We know that platinum sensitivity can be utilized as a biomarker for PARP response. And so if you give bevacizumab in the upfront, and the patient ends up having a platinum-sensitive recurrence or a long platinum-free interval, you potentially could give PARP inhibitors that next time around. Now, what's the pro for utilizing norepirib based on the PRIMA indication? Well, it's oral. So they don't have to come in every three weeks for infusions. You know, if the patient already has risk factors for perforation, there's, you know, bowel obstruction or a lot of prior bowel issues like IBD, certainly that could be considered as a pro for utilizing a PARP inhibitor. You know, one thing we talk to our patients about all the time is the HR deficiency tests that we have are okay, but they're not perfect. And we know they miss some patients. And so perhaps the patient in front of you that has a homologous recombination proficient tumor may be one of the ones who could potentially benefit from PARP. So there's no way to know that for sure. And in a patient that wants to do something and has heard about these PARP inhibitors and is really excited about it, that could be very reasonable. And I hear this a lot and I, you know, I go back and forth about whether I truly believe it, but this may be your only chance to give a PARP inhibitor. And yes, you know, do we believe that all patients in front of us should get a PARP inhibitor at some point during their disease course and Bevacizumab at some point? Potentially, these are two very active drugs. And so if you don't give PARP in this upfront setting, that patient may never be able to get it in the future. You know, we have biomarkers and we need to use them and it's exciting to have them. So I don't get super excited about that argument. And I think that we've got a lot of really exciting combination data that are coming down the pike. And so hopefully those will be the opportunities for those patients. And again, not necessarily for this case I've already started Bevacizumab, but if you haven't started Bevacizumab in the upfront setting, you can also consider active surveillance. So my, I made a decision because I'm being definitive and I am going to continue this patient on Bevacizumab for 15 more cycles. But I will just sum up by, we have to do better here. We are not getting as good outcomes as we need to be getting for these patients with homologous recombination proficient tumors. So please continue to participate in all the clinical trials that are ongoing. So we can start really impacting those patients, this big population of patients as well. And with that, I would thank you all for your attention and I'm going to stop sharing. Awesome. Thank you so much, Shannon. Yeah, thanks for following the instructions. I know Dave, I have to send them a couple, I know, I know, I have to definitely send definitely multiple reminders to stay on script. No, that's okay. And thanks so much for your presentation. We have questions coming in. So I encourage the audience, please submit your questions through the Q&A tab. They are showing up here on our side so we can see them in real time. So please send them in for all three of the panelists. And so I'll introduce the last of these clinical vignettes. Let me share my screen here. Alrighty. This is the problem with these things. We have so many slides to pick from. Well, I'll just do this. Sorry, I'm getting there. I have to share the whole screen here because I can't find the slide presentation, but I will present this. Okay. So the third case is a 63-year-old woman with a history of stage 3C high-grade serosevere cancer. Surgery, she had a near complete gross resection, left with kind of diffuse miliary disease. She had a panel testing done which showed no pathologic genomic alterations. So she was BRCA wild type, but the tumor tissue wasn't available for HR, particularly for HRD testing. Her CON25 at the time was 367. She had adjuvant treatment with Pac-Pax chiroplatin for six cycles. The CA125 normalized quite rapidly. Her exam was negative. Imaging was negative except for some post-op changes. And her CON25 was 13 at the completion of her six cycles of chemotherapy. And she did well for a year. And then her CON25 dropped from 18 at the previous determination to 164, which prompted a CT scan. And so the CAT scan, you can see on the right, the axial images of the pelvis. You can see there's a little bit of fluid and some nodularity along the anterior abdominal wall. She has ultimately about three to four, one and a one and a half centimeter densities in the peritoneum that were consistent with recurrent disease. So I'm going to ask Roz Glaspole to take this case on for a presentation of treatment for her recurrence. And I'll stop sharing and turn it over to Roz. That's great. Thank you very much. Just as I bring up my slides, just like to thank you all and apologize because I am not in my normal office and somebody seems to have decided to start cutting down a tree outside. So if it really gets really noisy, you might have to just cut me off altogether. What I just need to do is put this onto, just kind of get it onto the slide presentation, which is currently not quite working. There you go. On top, look on the top. There you go. You got it. Should be able to do it. There we go. Okay. So those are just my disclosures. So yes, Rob's described the case. So what next for this lady? So it used to be easy when really our only decision was when to start chemotherapy. But now we have this whole array of different therapeutic options we need to think about. So I just thought I'd take you through step-by-step the kind of way I would make a decision. So I think the first question, so she's relapsed greater than 12 months from her last platinum. So the first decision is, should we be doing surgery? And in the interest of time, I'm not gonna go into this in great detail. I'm not a surgeon and I know there was a great debate yesterday, but I think essentially the summary is that if there is an expectation that complete debulking could be achieved, then she should be having further surgery. We're gonna take from this one though, that she, if we go by the AGO criteria, she was not completely debulked initially. So I'm gonna sort of be brave as a non-surgeon and say, she's probably in my center and he would not be selected for surgery. So the next question is chemotherapy. And we know that the probability of responding to further platinum is a continuum of probability with increasing chance of responding with increasing platinum free interval. There is no magic change between five months and seven months. It is a continuum. And this was the data from BlackBridge back in 1989. And I think it's important to realize that this actually predated the time of CT scans. So this was symptomatic relapse. So this is people with rising CO125 triggering scans. So really we should be considering platinum again for people who've had a good initial response, even if it's beyond six months. But for our patients on this occasion, it's really a no brainer because she has had a 12 month platinum free interval. So she definitely needs platinum-based chemotherapy. And I'm assuming that she's reasonably fit. And I would therefore be offering her a combination therapy because we have the meta-analysis of trials showing a benefit with combination therapy, but it does come at the expense of greater toxicity. And I would probably choose carboplatin and PLD based on the CALYPSO study, which showed a small benefit in this case. And it always has the advantage of not losing your hair, which is often important. I've put here plus or minus bethasamab and it didn't say anything about whether she had bethasamab up front. So I guess we can assume she didn't. I'm going to come back to a little bit more about antiangiogenics later. I actually have quite an easy decision because in Scotland, it's not reimbursed in the platinum sensitive setting, but it'd be interesting to hear what other panelists say in their areas. So assuming she goes on to have a chemotherapy, if she has no response, that would be a really poor prognostic factor. And unfortunately we've been looking at non-platinum agents in clinical trials. But if she has a response, then the next question is the maintenance PARP inhibitors. And we've got several studies now. So the ARIEL3 with recaparib, NOVA with niraparib, and study 19 with olaparib looked at all comers and obviously SOLO2 again with olaparib was just confined to the brachymutant patients. And I think the striking thing is the consistency, despite differences in design, of the efficacy really across, not just the intention to treat population, but rather across the intention to treat population with the greatest benefit being with the brachymutant patients. But we know our patient didn't have a germline BRCA mutation. And I should have said at the beginning, although we don't have availability of full HRD testing in my center, we do have reimbursement for somatic BRCA. And that's something that I would get tested for her because that would be an additional piece of information that she could have a somatic mutation, even though she's a germline wild type BRCA. But what we do see is this significant benefit in BRCA wild type. Had we had the additional information of an HRD test, well, then we might be able to tell her that she would be more likely to benefit. But as in the first line setting, we do see a benefit in so-called HR-proficient tumors. So we can't really distinguish on the basis of a molecular test, are there any clinical groups that may or may not benefit? And really the answer is no. So in the analysis of NOVA and ARIL3, we've looked at complete and partial responses and not seen any significant difference in magnitude of benefit. Although I think it is important to note that both those studies did include patients only who had either normalization or near normalization of their CO125 as well as a partial response. So they really are really good partial responses and the license is broader and will allow any response. And we don't know whether those groups get the same benefit. In terms of age, not surprisingly, if you look at the older population, you have a lower incidence of BRCA mutant patients, but within their groups of mutation or wild type, they benefit similarly. And again, with platinum-free interval, although we see improving prognosis with improving progression-free interval, we don't see a difference in the magnitude of benefit from the PARP. So we don't really have a way to say that we would not give anyone a PARP. So what does this mean for patients? Because I've been talking about median progression-free survival, and obviously that is not necessarily a meaningful thing. What does it mean to patients? And unfortunately, it comes at the expense of toxicity. And those can be significant with really quite significant rates of grade three or four toxicities, but importantly also grade one and two, which in a maintenance setting can have a big impact. But what I do say to patients is even if we don't see that long a benefit as a median for the BRCA wild type, we do see long-term responders. And I think if you say to someone that you might have a two out of 10 chance of a durable response, then they may be prepared to take those odds. We also see improvement to the time to first subsequent treatment. And I think that's a probably more meaningful thing to a patient is how long is it gonna be before I have to have chemotherapy again? And also we see that the benefit is maintained beyond that first progression and into second progression. And we see a benefit in quality adjusted PFS and also this thing called QTOS, which is time without symptoms and without toxicity. So that's trying to get a kind of balance of what are the payoffs between improving progression and the type of toxicity. And so this is just some of that sort of visually demonstrated for the aerial three study, I think it was. And you can see that the toxicity, which is shown in the colors here, tends to, sorry, this is both NOVA and aerial three, but it tends to be early on. And once you've managed to adjust doses and adjust things, then that does decrease. And you do get the significant benefit in terms of time without symptoms or toxicity. And the really important thing for people, obviously, is overall survival. Now, we don't have the overall survival data yet for aerial three and NOVA, but we do for solo two. So this is the best prognostic group, but it was nevertheless really exciting to see this extension in overall survival with the addition of Olaparib in solo two. Didn't quite meet statistical significance, but if you then, in a pre-planned analysis, they looked at an adjusted population, adjusted for subsequent park inhibitor therapy, and that was statistically significant. And again, this important thing of time to first subsequent therapy with 28% of patients still alive and having not needed to receive a subsequent therapy compared to 13% in the placebo. So even when we're talking about recurrent disease, this really is a game changer. But so far for our patients, she wasn't in that group because she didn't have, we don't know that she has a somatic mutation and she is wild type in the germline. So I've mentioned the HRD tests and what does that really mean? I won't say too much because I think Professor Gordon is going to talk about this later, but essentially the tests that we have at the moment are measures of genomic scarring. There's a Myriad test and the Foundation test. So Myriad used in NOVA and Foundation test was used in Arial 3. I won't go into the details, but they're both predictive of response to PARP with a greater benefit than those with what's called HRD positive according to those tests. But they do have limitations because they are historical and static. So they're measuring the scarring that has occurred usually in a biopsy taken at presentation. But if there's a revertant mutation, say that scarring isn't going to disappear. So you're not going to see a change as resistance develops. They're also, we have the problem of heterogeneity as we know that hybrid serous disease is very, very heterogeneous. And so you only have a biopsy in one side. And the main problem is that they lack a negative predictive value as I said. So they don't exclude, they don't tell us who we shouldn't give a PARP to. And what we don't know is whether that is because they have to have a cutoff and we're therefore just not including some patients that do have functional HRD within their tumor, or whether it's actually that there's a different effect of PARP inhibitor in those HR proficient tumors that's producing the benefit. So there are a number of different approaches underway. The challenge for all of them is that they need to be validated in prospective trials. And I think it's really, really important and I'm very pleased to see the academic collaboration that's going on by Gineco where different approaches will be tested in those parallel samples. So just back to Bevacizumab, just as in the first line, we have good data in the relapse setting showing that Bevacizumab does produce a modest but significant benefit. And the recent AGO study showed that Bevacizumab can be combined with the copper pattern PLD regimen. And had she had Bev in the first line, the MITO16 study also shows that having Bev in the relapse setting will also can produce a similar benefit. And so if you compare those modest benefits with Bevacizumab to the benefits of a PARP in the HR proficient group, we're talking about very similar levels, but obviously what we don't have is the benefit of Bev in these different molecular groups because those studies didn't include that. So we don't know exactly what the benefit of Bev in that HR proficient group would have been. However, we will talk about that a little bit more. So what happens if your patient says, but I don't want to have chemotherapy, do I really need to have chemotherapy? So what's the evidence for PARP as treatment? And there are a number of studies that have looked at PARP as treatment and demonstrated that they do have efficacy, but the efficacy really is in the BRCA, the good efficacy is those BRCA mutations. But how about if we add an anti-angiogenic therapy and the Joyce Liu study with Olaparib versus Olaparib and Esidirinib and the Ziavanova study, looking at the addition of Bev to Olaparib. And we see that, yes, there is a benefit by adding an anti-angiogenic to a PARP as treatment. And interestingly, in this occasion, although the overall prognosis is worse for BRCA wild type, the added benefit of the anti-angiogenic is bigger in the BRCA wild type group. The NRG studies, this is slightly different because this was compared to looking at chemotherapy. And what we see is that, again, you do see a better effect with the combination of Olaparib and Esidirinib. Oops, sorry. But in our patient who was BRCA wild type, really chemotherapy would probably still be a better option for her. What we don't know is what the combination of an anti-angiogenic in the maintenance setting is. And we have a study running in the UK, which, and in other countries around the world, which will address this question. And that's the ICON9 study, which is looking at Olaparib plus Esidirinib versus Olaparib alone in the maintenance setting after a response. So just very quickly, what about the immune checkpoint inhibitors? The Mediola study was obviously an interesting study. It was a single arm study. It showed what looked like impressive results in BRCA mutant patients, but we know Olaparib is a good drug on its own there. So really we need to wait for the randomized studies, although this is obviously an exciting area. So questions for the future, we desperately need, we've already said it, we need it in the vedicism. In the first line setting, we need better positive and predicted markers, and I think that's gonna require an understanding of the resistance mechanisms. We will be looking at other PARP, or are looking at other PARP combinations, not just with antiangiogenics, and that's a really exciting area. And there will be this increasing question of PARP after PARP as we move the PARPs into the first line. But going back to my patient, so my recommendation, I'm gonna jump in and say, I'm not gonna operate because she wasn't completely debarked beforehand. I know that may be argued differently, but we'll assume that it was a very good surgeon at the beginning that didn't manage to debark her first time around. I'm gonna say she has platinum-based combination chemotherapy, followed by either the ICON9 trial, if she's eligible for that, or one of the other PARP inhibitors. And I would do that regardless of her BRCA mutation status or HRD status, and I personally would reserve the vedicism app at the moment for platinum-resistant relapse if I was not able to offer ICON9. I think the one thing is that we do live in very challenging and difficult times now within the COVID-19 pandemic, and certainly at the height of the peak, we did consider, and we have now got approval and funding for within the UK, a PARP inhibitor treatment for those who have got a BRCA mutation instead of chemotherapy as a way of reducing their needs to visit the hospital. And that's all. Great, thank you so much, Raz. We have a couple questions coming through, so I'll address them quickly. We have a few minutes for Q&A, and I wanna make sure that we get to some of these. And so the first question, Shannon, this came from two people, and it's basically the same thing. It's addressing this patient who was young and kind of had an emergent surgery, and we run into this type of issue all the time. The question is, let's say you evaluated her and felt that she should have that mass resected so that she was essentially a complete gross resection. Let's just take that as the example. How would that change your management of her? So again, speaking about the adjuvant and the maintenance setting. Yeah, it's a great question, and it is definitely a direction I could have gone in. I chose to kind of go through like, well, she just had a massive bleed and we're just gonna get her on chemo, and then she responded, and there was no residual disease to resect, but the discussants are asking really good questions. So if I was able to take her back to surgery, got her completely debulked, then I would likely just do paclitaxel carboplatin every three weeks kind of routine. And then you have a decision in front of you, which I kind of alluded to in the other considerations piece of, do you give that patient a PARP inhibitor or not? Young patient that hopefully was very healthy and would wanna do something, I would definitely consider norepinephrine maintenance for her because if she is one of the patients that benefits despite her tumor being tested as homologous recombination proficient, then we have a real win here. And I'd rather take the chance of the toxicity to try to get that benefit. But this is where shared decision-making is so important, and you really have to get a sense of where your patient is at. We all have these patients that wanna do something. We also all have patients that are like, I'm done if you can't really offer me a huge benefit in survival or guarantee me a huge benefit in survival, I'd rather just wait and watch. And so I think those are the two benefits, but it really gets at that point I was making that we don't really have great options right now for homologous recombination proficient patients. Yeah, and I think it has a little tie-in to Charlie's talk about HRP or the HR assessments because I always wonder whether or not the benefits that we're seeing in the HRP cases are just bad testing and not really efficacy apart. So, but I'm curious to see what he says about that. So Roz, a question that came in just to address, we appreciate, I think, the global differences of availability of drugs, and I know that's influenced your decision-making, and thank you for focusing on what you would really do in real time. The question is, in your patient, who you gave a discussion about the use of Bevacizumab, the question is, is there, let's say that you got a really good response, you added Bevacizumab because she was platinum-sensitive ascites, and you got a good response. At the consideration of the maintenance place, your options would be to continue, Bev, as you shared with all of those trials, Oceans 213 and 221, you know, that showed a benefit, or switch over to one of the, you know, Naraparib, Rucaparib, Elaparibs, or both. So assuming you had access to everything and you believed all the data, what would you do? So I think, I mean, at the minute, we don't have the data for the combination in the maintenance setting, but at the same time, it's hard to see that that would be different from the treatment trials. So I suppose if you'd seen a really good response and you were on both, I would be tempted to be carrying on with both. You could also argue, though, that if you've seen a really good response, then that could be the platinum that's giving you the really good response. And the patients that I probably worry about the most are the ones that don't have a very good response. And you know that that response is gonna be short, and I'm not sure that those patients, we don't have the data really for them with PARPs because the studies really did select those with really good either complete or partial responses plus complete CO125 responses. And in reality, we see a lot of patients who still have disease, you know, they have a bit of response, CO125 comes down a bit, but it's not great. And those are the ones where if I had started the bed from the beginning, because I was able to do that, then you would think I'd actually quite like to keep it going because they may actually be getting maintenance of that response for longer. But yeah, it's a real challenge. And as I said, I have that easy option because it's taken away from me. But I think my bias is that probably those ones that are not having a good response are the ones that are then gonna benefit best from the combination of PARP and anti-angiogenic. And that would be really nice to be able to do that. But the trouble is, is that we keep adding different agents and new things coming in. We can't keep giving everything to everybody. So there has to be at some point, we're gonna really get stuck. And we're already stuck because of funding, but we'll be stuck with toxicities as well. So we need our translational teams to really get them in terms of the markers that are gonna help us to decide what to do. Okay, great. Thank you so much. And Dave, just a quick question for you, because I can't give you a softball, but, you know, in your patient, you know, she, you know, looked like she responded and you provided us all those choices. What would you do if the patient, you know, didn't really have a complete response at the completion of her chemo, Bev strategy that you talked about? You're on mute again. Why is he so bad at this? I don't know. Can you hear me? He's using his lab coat in his attic. Well, and I, in my big old butt, when I lean back, I accidentally hit this microphone thing. That's really the issue. I'm too fat. So, you know, let's be honest here. Shannon, you don't have to shake your head yes. I didn't say anything. I was mute. I was. So, you know, I, you know, I think the partial response patient, you know, what we used to call that platinum refractory, now we call it going to maintenance. You know, I'm not sure when we changed that, but, you know, that patient obviously is extremely challenging. In that population, you know, do you do two more cycles of chemotherapy? If she continues to have a CA 125 response, I think it's a very reasonable option. And I would definitely consider that. I also would look to use PARP and Bev as a treatment. You know, we've seen some modest efficacy results, even in HRP patients. We're going to see more in a combination here at Esmo. So I think, you know, we're really looking at utilizing PARP in that scenario, like we did in some of the aerial trials as a treatment modality, not truly as a maintenance. And so much of our, you know, we call it maintenance, but we all know it's a treatment. Yeah, great. Okay, thank you so much. Doug, I'm going to pass it back over to you to go on to the next session. Thank you for all my speakers. Great job, you guys. Thank you to everyone. That was a wonderful session. It really highlights, you know, the way that molecular medicine has driven the use of our drugs to date, and also the unmet needs that we have in taking molecular medicine to the next level, which is why we have two closing talks that will discuss these issues and these particular unmet needs. And if you're just tuning in, you're watching the master session on ovarian cancer on IGCS channel number one. We have two more talks, and we'll now hear from Dr. Charlie Gourley about molecular testing for initial therapy beyond BRCA testing. And Dr. Gourley, please go ahead. You're sharing your screen, but if you click on the top display button, you can make it go into one screen instead of two. And there's a little box that's blocking part of your screen, probably from something else that's open. But if you go up to that middle thing, display something or other, display settings, and choose the other one, your slides will look a little better. Sorry about this. I don't seem to be able to get onto that. Yeah, go ahead and click. Click that thing. Yeah, click that. All right, now go up to the top where it says display and settings. It's a little more, yeah, there you go. Yeah, click the other one. Yep, you got it, yes. Okay, if anyone else needs IT help, please contact Dr. Coleman. What I'm here for. Apologies for my IT ignorance. Okay, thanks very much. Here's my disclosures. So we're talking about molecular testing for initial therapy beyond BRCA. I'm going to quickly talk about the molecular subgroups of high-grade serious ovarian cancer. And really what I'm talking about here is just high-grade serious ovarian cancer. There's not enough time to talk about other subtypes. And I'm going to talk about which molecular lesions are actionable now, aspirational aims for the future, and then finish with a quick case vignette. So the molecular subgroups of high-grade and serious ovarian cancer. This is a picture many of you have seen before or variants of which many of you have seen before where you've got one half of the pie chart having the various abnormalities that cause homologous recombinational repair deficiency, all the germline and somatic BRCA abnormalities and the other abnormalities in HRR genes. And then on the left-hand side of that pie chart, you've got the very interesting sort of under-researched abnormalities such as cyclin E amplification and RV1 loss and NF1 loss and PTEN loss. I'm going to talk a lot more about how we can maybe improve this pie chart later on, but just to mention that there are other things, some of which I'm going to talk about, obviously HRD scores I'm going to talk about, immune scores, Brad Nelson hopefully will be talking about various immune markers, so I'll mark that from that. And then epigenetic changes are important, but they've not really been incorporated into many of our tests until now. And then there are these other things, mutational signatures, that's the pattern of mutations that happen in tumors, or structural variant signatures. Structural variants are large-scale deletions, inversions, translocations. And there's also this thing called whole genome duplication that happens in about half of our high-grade serious ovarian cancers. And at this point in time, we don't really know how to incorporate that. Some of the HRD tests incorporate these things, but many of them, including the ones we use in ovarian cancer, do not. So what molecular lesions are actionable now in the first line setting? I think it's fair to say that HR gene mutations are actionable, they certainly give us some information. And David O'Malley spoke about this in the Arial 3 study. This slide is from study 19, which was one of the first PARC inhibitor studies, and it showed that 22% of the BRCA wild-type patients had HRR mutations. And certainly in terms of progression pre-survival, the possession of an HRR mutation suggested you were more likely to benefit from a laparotomy than if you were HRR wild-type, if you were BRCA wild-type. So I think this adds information. And I think the only other test right now in the first line setting that adds information are these HRD scanning assays. And many of you will be familiar with these. There's Myriad Mitrois, Foundation Medicine LOH assay. There are actually other HRD scanning assays. There's a test called HR-Detect, which was developed in breast cancer and is a bit more sophisticated. And these are okay, but they have their limitations. If we think about how predictive these HRD assays are, and I've listed here a couple of the first-line studies, PRIMA and PAULA-1, and a couple of the relapse studies, NOVA and ARIO-3. These studies use as their predictors either Myriad Mitrois or the Foundation Medicine LOH assay. And what is common amongst all these studies is they show that HRD tests can enrich for patients more likely to respond to PARP inhibitors, as shown by the fact that the HRD-positive patients, the hazard ratio is lower than the HRD-negative patients in each trial. However, it's only in one study that the HRD test identifies patients unlikely to respond, and that study is PAULA-1. It's quite difficult to explain this, and I know a lot of people have racked their brains. My feeling is that the most likely explanation for this is that the other three studies here all have a fairly high level of clinical pre-selection for HRD, i.e., in the relapse disease studies, it was only patients that were platinum-sensitive to more than one line of platinum. And even in PRIMA, there had to be a signal of platinum sensitivity of some sort for the patient to get onto the study. So under those circumstances, when you've got clinical pre-selection, the HRD test has to work a lot harder in order to pull out those patients that don't benefit or exclude those patients that don't benefit, whereas probably in PAULA-1, it's a bit more of an unselected population, so it doesn't have to work so hard. And as has been mentioned, Shannon Wesson mentioned this, HRD tests, they are far from perfect. So there's false positivity because there are patients who have a HRD-positive score but have got reversion mutations, for example, and have actually got restored HRD, but they will still test positive. And then there's clearly a false negative element as well. So what about the aspirational aims for molecular testing going forward? I showed you this figure earlier, but I think there's a lot of mileage into looking more at mutational signatures, structural variance, and whole genome duplication in high-grade serious ovarian cancer. And working out what these things actually mean. Also, let's take this little figure on the left-hand side, this pie chart we see a lot. Well, it is a simplification, it's an oversimplification. We've tried to add a little bit of depth, and we've made this thing that we called an onion plot, and I'll go through it in a little bit of detail so it's a little bit more interpretable for you. So inside of the onion, you've got P53 mutations, and pretty much all high-grade serious ovarian cancers are abnormal in the P53 pathway for one reason or another. Then in the next layer, similar to the previous pie charts you've seen, we've got on the right-hand side all HRD mutations, and on the left-hand side, the HR-proficient aberrations. The yellow bar is cyclonealtification. This is a really important group. And then the orange bar is other HR-proficient abnormalities. And then if I add the last bit of the onion in, you can see that the other HR proficiency is due to P10 loss, NF1 loss, or RB1 loss. Now, it used to be thought that all of these things were mutually exclusive from the HRD abnormalities, but it turns out that you can have an HRD abnormality and still have NF1 loss and still have RB1 loss. And actually, that might be important. Anna DeFazio's group in Australia looked at the patients who had HRD loss and also had RB1 loss. And when she looked at exceptional responders in high-grade serious ovarian cancer, most of those had HR deficiency. But she showed that loss of RB1 protein significantly predicted that the patients would do even better than the patients who were HRD but had retained RB1 protein. So there's clearly additional areas or additional levels of complexity here that we don't understand. There are also subgroups that we need to know more about. The homologous recombination proficient group is clearly a massive area of need. So hopefully, we can pick through these. So just for the last minute, I'm going to go through a very quick case. A 54-year-old woman who presented in a familiar way with abdominal discomfort and distension. CT scans showed a seven-centimeter anexal mass, a mental and peritoneal disease. She underwent primary debulking and was debulked to zero macroscopic residual. And she had a high-grade serious ovarian cancer, stage IIIc, T53 aberrant. In Scotland, we do jam-line panel testing, and we found that she had a RAD51C mutation, which is obviously one of the HRR genes. She got six cycles of adjuvant carboplatin and paclitaxel. And so the question is, would you give her maintenance parthenometer therapy in this first-line setting? Now, of course, most of us would say she's got a high chance of benefiting from a parthenometer. If you live in the U.S. now, you do have licensed options. You could potentially give her an Oraparib or Olaparib plus Bevacizumab. Elsewhere in the world, you'd have to fight her case. So what we would do is we would put in an individual treatment request to access one of the parthenometers. And we'd probably provide evidence on the basis of the relapsed disease studies that show RAD51C mutations, predict sensitivity to these drugs, and kind of dovetail that into the solo RAD51C. So that's really probably it for me. I think I'm probably short on time, so I'll skip the summary. I'll just say thank you very much for your attention. Great, that was a really wonderful talk, Dr. Gorley. And now the last talk of the master session on ovarian cancer. And thank you to all the 227 people who are still with us. Will be a wonderful talk from Dr. Brad Nelson on biomarkers for immunotherapy response. Dr. Nelson, please share your screen as you're doing, and go ahead and begin. Thanks very much. Well, good morning, afternoon, evening, everybody. And thanks to the organizers for inviting me to present today. Is this showing up okay? Yes, looks good. Great, thank you. So I'm gonna just give a very brief overview of what I consider to be the latest developments in immunotherapy biomarkers, and really focus my discussion on high-grade serous ovarian cancer. So I have nothing to disclose. And I'll just start by, the problem that we wanna solve and build upon is the fact that ovarian cancer in general is not responding well to today's form of immunotherapy with a blockade of the PD-1 and PD-L1 pathways. Unfortunately, the objective response rates are in sort of the 10 to 15% range in contrast to other malignancies where these agents are showing much more efficacy. So how can we understand this and how can we build on this? So I think one of the issues with high-grade serous carcinoma and other malignancies as well, some other malignancies, is that in this type of cancer, the PD-L1 expression as shown here with the brown dab staining is primarily found on the macrophages in the tumor rather than on the tumor cells themselves. So with this three-color staining, we're showing the tumor cells as WT1 positive in blue and the macs are in red and PD-L1 again in brown. And so you can see there's a lot of PD-L1 expression in these mac-rich stromal regions and much less so in the malignant epithelium with some exceptions as shown here. So it could be that when we're unleashing, the giving patients these agents, anti-PD-1 or anti-PD-L1, what we're primarily doing is unleashing a response against tumor-associated macrophages rather than the tumor cells themselves. So how do we go forward from here? This is an overview of all the different checkpoint pathways, the receptors found on T cells in the bottom here and then their counter structures found on the target cell above. So a very complex molecular dialogue between T cells and tumor cells. The T cell receptors here, PD-1, PD-L1 are shown here. How do we rationally choose among these other possible checkpoint pathways and try and find ones that might be more efficacious in the HGSC setting? So obviously this is a huge area of research and I've only got 10 minutes. So I'll just tell you one little vignette from some work coming out of my lab. The first part of it was published last year in PNAS where we did a pan-cancer analysis looking at TCGA data. So these are all the different types of cancer. And what I'm showing you here is a heat map of the association of different immunologically actionable genes or proteins and their association then with the cold tumor phenotype across these different types of cancer. And there's a number of interesting things that came out of this, but what I'll just tell you about is that we followed up on is this gene, a PDR, a poliovirus receptor, also known as CD155, which again shows a strong association with the cold tumor phenotype across cancers, including ovarian cancer. So we worked up a multicolor immunofluorescence stain for this. So CD155 is what I'll be talking about now. Again, this poliovirus receptor and applied this six-color stain then to a series of HGSC cases. And this was published just recently in Gynae Oncology. And what's really interesting and I thought therefore worth just pointing out is the markedly different expression pattern we see with CD155 versus PD-L1. So again, PD-L1 is in this sort of turquoise color, mostly lighting up the macrophages in the stroma of the tumor, whereas CD155 is lighting up the tumor cells themselves in this orange color that you can see here. So really quite contrasting and complementary expression patterns. And as predicted by our bioinformatic analysis that led us to even be looking at this, what we find when we look at the association between PD-L1 on the left or CD155 on the right, if we look at expression in cold tumors, these are all HGSC tumors, versus hot or versus what I call the warm tumors. These are ones where the till are in the stroma of the tumor, but not the epithelium. We can see as has been reported by us and others before that PD-L1 is mainly found in the hot tumors in those macrophage rich regions. And interestingly, CD155 is found in hot tumors as well, but even more so in cold tumors. So suggesting this then as a relatively novel checkpoint one could consider. I think I may have forgotten to mention, CD155 is the ligand for TIGIT. TIGIT is the structure found on T cells, the inhibitory molecule. So CD155 and TIGIT form, if you will, another PD-L1 like pair. And there are TIGIT, anti-TIGIT antibodies in clinical trials. And we look forward to those results to see if, of course, hopefully in this setting, we'd get a more efficacious response. So the first point that I've tried to make is that there may be more important checkpoints than PD-1, PD-L1 in high grade serous carcinoma. I've talked about the TIGIT CD155 pathway. And of course, there may be others. And this is fortunately a very active area of research. The second point I wanna make is that, what we've learned in the last few years is that when we look at TIL in ovarian cancer, yes, they are highly prognostic. Many groups have shown this now, including our own. But when we look in detail at them, it turns out that the majority of TIL don't seem to recognize the tumor. And this is a study from Tom Schumacher's lab that our group contributed to as well, where individual T cell receptors were pulled out of TIL from, and I'm showing you an HGSE case, and each T cell receptor was tested for recognition of autologous tumor. And of the 20 or so TCRs that were tested here and shown here, only one of them actually reacted to the tumor digest. The rest did not. And not shown here is that some of these we were able to show actually recognized viral antigens, CMV, EVV, influenza, and so on. So, and we call these bystander T cells. And this is kind of how the immune system works. Whenever there's inflammation, all the T cells go rushing to that site, including irrelevant T cells or bystander T cells. Presumably, they're looking to see if their antigen is there. But the summary statement then is that TIL are highly prognostic, but somewhat paradoxically, only a small proportion of them seem to actually recognize the tumor. So the race is on then to find better markers going beyond just staining for CD8, for example. Can we identify the bonafide tumor reactive TIL among all the bystanders that are in the tumor? And so this is work that, from my lab, building on work from other groups in other malignancies, where they've nominated a few other markers as potential markers for the tumor reactive T cells. So CD39, PD1, and CD103, as a triple combination, the hypothesis is that these define the tumor reactive subset. So we worked up a multicolor panel, including these markers, as well as CD8, and looked at a number of tumors, and you can see the sort of jellybean-like patterns that you get when you look at this number of markers, in CD8 TIL. But we do indeed find a fairly rare population of cells, and here's an example here, that are CD8 positive, and also express all three of these other markers. And for the sake of time, I won't say much more about them, other than when we look at their prognostic significance, we find they are strongly associated with survival. So that's the top line here, are patients with these triple positive CD8 positive TIL, and the blue line is any other combination of these markers among CD8 positive TIL, or putting them below median. And this is restricted to patients who are all TIL positive. So these are all, on both lines, these are patients who have hot tumors that have CD8 positive TIL, but we're now further stratifying on these markers. So giving us some confidence that we're moving in the right direction here, in coming up, coming to understand the phenotype of the most important TIL. So that's my second point, is that we need new biomarkers to identify the tumor reactive T cells. And I've talked about CD39, PD1, and CD103, as attractive candidates at present, and then others are also under investigation. Then the final point I'd like to make, is that it's not just about T cells. Of course, they get a lot of attention, they deserve a lot of attention, but they don't work alone. And so my group for a number of years now has been looking at the role of B cells and plasma cells in the anti-tumor immune response. And this is just an image from a very hot HGSE tumor, where we've done staining for all these cell types. And you can see in regions that have a lot of T cells, which are the red cells. In a really hot tumor, you also have shown in green B cells and plasma cells sprinkled among the T cells. And then in blue, we're showing the macrophages. And in fact, this is where we see the PD-L1 lighting up, is in these real hot zones where all of these cell types come together, and we think are working together to destroy the tumor tissue. So this is unpublished work, part of a DOD-funded consortium called MOCOG that is really built on the AUDA-OCAC consortium and is led by Lee Pierce and Malcolm Pike, and with many other great investigators to work with on this. And we're looking at the determinants of long-term survival in HGSE, using a very large international cohort of short, medium, and long-term survivors. And this is just interim data at this point, again, unpublished, but showing that we do find as expected that long-term survivors of ovarian cancer have an enrichment for epithelial CD8 positive T cells. That's what we expected to see. But the new finding coming out of this is that the presence of stromal plasma cells is also highly enriched in the long-term survivors. So supporting the idea that these B cells and plasma cells are good and that they are contributing positively to antitumor immunity. And this concept is really gathering steam in a number of other tumor types as well. There were three papers published together in Nature earlier this year, looking at the B cell response or related tertiary lymphoid structures and positive associations with survival and immunotherapy response in melanoma and sarcoma. So we see this as a new area, relatively new area with lots of opportunity. So that's my third point, is that tumor infiltrating B cells and plasma cells are strongly associated with survival in HGSE and other malignancies. And we think this opens up a whole bunch of new opportunities, targets for immunotherapy. And I will stop there and again, thank the audience and thank the organizers for inviting me. Great, thank you, Dr. Nelson. That was a wonderful session. You can stop sharing your screen if you want and we'll move on to a brief panel discussion before we close up. We have a lot of questions coming in, but you still have time to send your questions into the Q&A box. And since Dr. Nelson just finished, we'll start with one or two questions for him. You elegantly showed that it's not just the T cells and I think you use the term that some T cells are bystanders in the tumor microenvironment, yet as we date back 20 years almost, we know that if you have tumor infiltrated lymphocytes slash T cells, you do better. So are we talking about a subset of a subset here and could you make any comments as to like tumors with bystander T cells versus maybe activated T cells versus those that don't have as many bystander T cells hanging around waiting for the bus? Right, so on that, the question of bystanders, the problem is we don't, I've given you some early evidence now that we think we're getting, finding some good constellation of markers to separate the bystanders from the real good guys. But this is all early stages. So we lack really good, reliable, well-validated markers for the bona fide TIL. I think within the next few years, we're gonna really figure this out because we've got the tools, we've got the samples, we and others to figure this out. It's a lot of work to actually show for any given T cell that it actually recognizes the tumor. That's somebody going in the lab and doing a real experiment. So it can be done, but it's not as easy as just staining for marker. But with single cell sequencing approaches in particular, that's really helping to figure this out. And my prediction would be once we figure out these markers and I've talked about three that might seem to be pointing us in the right direction, that we're gonna, yes, we're gonna be starting to stratify patients, not just do they have CD8 TIL, but do they have the right kind of CD8 TIL? And then how does that then relate back to response to checkpoint blockade, whether it's PD1 or some other novel agent. The B cells, I've introduced them today as well. They're a subset of the CD8 positive tumors. So you don't see B cells off on their own with no CD8 TIL. They're within the hot CD8 positive tumors. They're a subgroup, and you can think of them perhaps as the super hot tumors that have both the T cells and the B cells. Well, I definitely understand that all this lab work is very hard and challenging. Just one more question for you. You showed nicely that a lot of the PD-L1 expression is on the macrophages. Is that unique to ovarian cancer, high-grade serious ovarian cancer compared to other solid tumors, or is that a common, broad phenomenon? You see it outside of ovaries, so you see it in breast cancer as well, for example, triple negative breast cancer and some other malignancies. It seems to be a subgroup of solid tumors where you have the predominantly macrophage expression of PD-L1. Lung cancer, and we've looked at lung cancer specimens side by side. That's where you can actually see really nice dark staining of the malignant cells themselves and quite widespread through the tumor. We don't typically see that pattern in HGSC, so it's an interesting difference in tumor biology there. Great, thank you. We have a couple of questions for Dr. Gourley. The first question is, you very nicely showed the molecular features of the tumors. I love the onion plot. It's very elegant. I'm going to probably steal that for you for future talks, but I think you were mostly talking about primary tumors. And so, we're not bad at treating primary ovarian cancer. Of course, you want to get better, but any comments on comparing the molecular features of primary tumors compared to recurrent tumors and what things might change, and is it important to re-biopsy patients to get the most current molecular characterization rather than one that was before chemotherapy was given? Yeah, my understanding from the data, and this was looked at a bit in some of the aerial studies where they had primary material and then material relapse. And certainly in terms of markers that would predict sensitivity to PARP inhibitors, it doesn't seem as though those change a lot, although there is some evidence of the epigenetic status changing. So, BRCA1 or RAD51C methylation state is changing and therefore that could influence PARP inhibitor sensitivity. But for most of the genomic features, of course, you could pick up secondary mutations. If there was a germline BRCA1 mutation, for example, you could pick up secondary mutations. So, that may be useful. More generally, there's a feeling that there's a drive towards this sort of NF1-activated, MAP kinase-activated tumor. And this is more, I guess, transcriptomically defined, but this sort of EMT signature, it's called, in the relapse setting. And so, if we ever get to a situation where we've got a handle on how to treat these MAP kinase-activated EMT tumors, then I think secondary biopsies will be even more important for that. Great, thank you. And another question for you. Would you give a PARP inhibitor to the patient that you discussed that had the RAD51C mutation? Oh, yes, I would try and get it. We would have to ask basically for it here, but I would want them to have a PARP inhibitor for sure. Great, and we'll just close with one last question for Dr. Nelson. Regarding biomarkers, you seemed to allude that single, you showed that single parameters are really not sufficient. What efforts are going on to develop a comprehensive panel of markers to better characterize the tumor immune environment? Yeah, good question. I think, I would say single-cell sequencing techniques are really gonna open this up with, for example, the 10X Genomics single-cell sequencing platform. We're able to get from individual T cells, both their molecular phenotype from the transcriptome as well as the matched T cell receptor, often beta chains, all that data in one experiment. So we can actually look at T cell phenotypes and line them up with their clonotypes, as we say, and then do experiments to see of those clonotypes, which ones actually recognize the tumor. So this has been a real groundbreaking technology for finally getting at this question of finding the needle in the haystack in terms of the most important T cells. Great, well, thank you very much. This has really been a wonderful session, and this channel will self-destruct in two minutes, and so I'll pass it over to Dr. Coleman for a 45-second closing. Thanks so much, Doug. Yeah, this is really a great session, and I'm really pleased, and thanks to all the speakers. This is fantastic. We covered a lot of information on the latest in ovarian cancer and diagnosis and treatment. So we're gonna close this session. Next on this channel coming up will be an industry sponsor, supposedly from Novacure, that will be addressing the role of tumor treatment fields in the management of recurrent ovarian cancer. Thank you all for tuning in for this master session. Make sure you explore the rest of the meeting portal when this program is done. For the day, there'll be a lot more to see. If you missed any of the part of this session or the others, these will be available on the meeting portal for the next 24 hours until September 24th. So with that, I'm gonna close it out, and once again, thank all my co-sponsors, co-hosts, and the speakers for their excellent job today. Thanks. Bye-bye.

debate

primary debulking surgery

neoadjuvant chemotherapy

ovarian cancer

complete gross resection

patient selection

surgery

improving survival

challenges

limitations

management

recurrent ovarian cancer

specific patient

PARP inhibitors

maintenance therapy