Welcome to Preparatory Session 2. I'm Dr. Matthew Powell from Washington University School of Medicine in St. Louis, Missouri. I really want to welcome you to the next, over the next 30 minutes, we'll be hearing three abstract presentations and then a brief distillation. Then we'll hear three more abstracts moderated by Dr. Renee Pereira and open up for a live session, question and answer session. It's a well-packed session, so stay tuned. And now we're going to move on to the first abstract. We'll hear from Dr. Emily Thompson to avoid any delays in the lengthy introductions. I'll turn the screen over to Dr. Thompson now and she'll present her abstract. Thank you very much for that introduction. I am delighted to be speaking to you today from Vancouver about our findings from a multi-center Canadian study evaluating variation in the practice and diagnosis and treatment of women with endometrial cancer in the era of molecular classification. I have nothing to disclose. You'll all be aware that histomorphologic classification of endometrial cancer is imprecise, especially in the setting of high-grade disease. By contrast, molecular classification enables objective and reproducible classification of tumors and provides prognostic and predictive information. Further, recommendations for molecular characterization have been integrated into the very recently published fifth edition of the WHO classification of female genital tumors. You'll be aware of pragmatic molecular classification strategies used to stratify endometrial cancer. The PROMIS molecular subtyping strategy stratifies endometrial cancer into four risk groups by using immunohistochemistry for p53 and mismatch repair gene expression, as well as targeted mutation testing for pathogenic polymutations. This strategy is able to pragmatically stratify cancers into four prognostically relevant risk groups. The objective in our study was to describe patterns of p53 and mismatch repair IHC use in a single calendar year and to describe the distribution of molecular subtypes of cancers within and across participating sites. We also aimed to evaluate the significance of unusual IHC findings. This slide depicts a map of participating sites. In black, we have cancer care centers distributed across the country, and in green, you'll see our community general hospitals within the province of British Columbia. From each site, we requested FFPE tumor tissue from either a biopsy or hysterectomy specimen and from every endometrial cancer diagnosed in 2016. All cancers underwent PROMIS molecular subtyping using the following iterative strategy, and clinical data for each case was collected using a standardized collection tool. To date, we have identified 1,453 endometrial cancers from 30 participating sites. Results presented today are taken from an analysis of the first 862 cases. These originate from seven major cancer care centers and seven community hospitals here in Canada. This figure depicts the distribution of molecular subtypes within and across the various endometrial histotypes. As you can see, endometrioid endometrial cancer shows the most variation in molecular subtype, with P53 wild type being the most commonly encountered subtype. In non-endometrioid cancers, P53 abnormal disease predominated nonetheless. Diversity in molecular subtypes was seen across all histotypes. This slide depicts the distribution of molecular subtypes in cancers across the various participating sites. I'd like to draw your attention first to site three as depicted on the x-axis. This is the amalgamation of the community general hospitals, and naturally, we saw fewer P53 abnormal cases in this cohort. In sites four and five, we saw relatively larger numbers of P53 abnormal cancers, and this is thought to be referral bias at these institutions. Conspicuously, the proportion of mismatched repair deficient cancers was relatively stable across all participating sites. This slide depicts the routine use of immunohistochemistry by participating site. Mismatched repair IHC was performed in 23.5% of cases overall, but its use varied from between 4% and 80% across participating sites. P53 IHC was performed in 15% of all cases, and its routine use varied by site from between 2% and 47%. Sites with lower rates of MMR IHC use had lower rates of hereditary cancer program referrals. Concerningly, 11.5% of cases recognized in 2016 as mismatched repair deficient did not undergo further evaluation using hypermethylation testing or via hereditary cancer programs. Switching gears now to unusual P53 IHC findings, we observed P53 abnormal cancers across all endometrial cancer histotypes, including low-grade endometrioid cancers. Subclonal P53 staining was encountered in 3.9% of cases and significantly associated with the presence of pathogenic polymutations. You may not have encountered the term subclonal in your pathology reports, but chances are you've seen a description of a cancer which shows an abrupt transition between wild-type staining as seen here on the right and abnormal staining as seen here on the left. Going forward, when you encounter these cases, I'd like you to consider advocating for polymutation testing in these patients in order to avoid misclassifying risk of these cancers. Other highlights, subclonal mismatched repair IHC expression was seen in 3.5% of cases, and we know from the literature that this most commonly occurs in the context of hypermethylated tumors. Mismatched repair deficiency significantly associated with the presence of lymphovascular space invasion and molecular subtypes significantly associated with clinical outcomes even in the setting of low-stage disease. In one participating site, we had the opportunity to evaluate the relationship between PROMIS molecular subtype and lymph node status. Here, all cancers were uniformly sampled by a full lymph node dissection. Here, we found that PROMIS subtype significantly associated with lymph node status, and this was due predominantly to the strong association between nodal metastases and p53 abnormal disease. In closing, we encountered significant variation in pathology practice and view this as representing opportunities for improvement as we move towards better quality care. Precedent in other cancers has occurred, and molecular subtyping is now routine in breast, colorectal, lymphoma, and malignant gliomas of the central nervous system. This means that for many of you, the infrastructure for molecular subtyping of endometrial cancer probably exists in the hospitals where you're working. It will be a question, though, of addressing how to invest in that infrastructure in order to accommodate the increase in use for endometrial cancers. As mismatch repair and p53 IHCs take on increasing recognition as biomarkers, there will be a need for standardized specimen handling protocols to optimize IHC interpretation. I strongly encourage you to work with your departments to achieve this. I'd like to acknowledge the contribution of our collaborators and my supervisors, of course. Lastly, I'd like to thank the Michael Smith Foundation for Health Research for their support of this study. Thank you so much. I'd be delighted to take your questions when the opportunity presents. Thank you, Dr. Thompson. Congratulations on your excellent work. That's fantastic, and thank you for staying on time. Now we hear from Dr. Seoyoung Rachel Kim. Dr. Kim is going to present her abstract now. Please upload your presentation. Thank you. All right. Thank you for that introduction. Are you able to see my slides? Yes. Perfect. Thank you for the opportunity to talk about clinical implication of mismatch repair deficiency in endometrioid endometrial cancer through a prospective study. I have no disclosures. As we know, DNA mismatch repair system maintains genomic integrity by correcting base pair substitution mismatches that are generated during DNA replication. The four main proteins that are functional in that machinery are MLH1, MSH2, PMS2, and MSH6. Deficiency in the mismatch repair system or loss of function in the MMR machinery can be due to different causes. Now in the literature, approximately 20 to 40% of all endometrial cancers are found to be MMR deficient. The most common cause is sporadic due to MLH1 promoter methylation. The remaining 3% to 6% are due to biallelic somatic mutations in the tumor or germline defects in one of the MMR genes, which we know as Lynch syndrome. Unlike in colorectal cancers where MMR deficiency is clearly associated with favorable prognosis and distinct clinical pathological characteristics, the relationship in endometrial cancers is not so clear. Some studies have reported significantly better survival among those women with MMR deficient tumor, whereas others have found unfavorable or no difference in their outcomes. A meta-analysis looking at mostly retrospective studies have found significant evidence of heterogeneity in the study population, histology type, and the methods of immunohistochemistry, limiting the conclusion about the relationship between MMR deficiency and the clinical outcomes. So the objectives of our study was to, number one, compare the MMR deficient and MMR intact endometrioid endometrial cancers with respect to their clinical pathological characteristics, survival outcomes, and their recurrence patterns, and to discriminate the differences in clinical outcomes according to their mechanism of mismatch repair loss. Now, this was a prospective court study from 2015 to 2018 of newly diagnosed endometrial cancers across three cancer centers in Ontario, Canada. All tumors were assessed for mismatch repair protein expression with immunohistochemistry, and all had clinical genetic testing if they met the appropriate criteria. We then compared the clinical pathologic details, survival outcomes, and recurrence patterns between the MMR deficient and MMR intact cohort. In terms of our study population, of 665 patients that consented, 492 had endometrioid endometrial cancers, and of those, 144 were MMR deficient and 348 were MMR intact. For the MMR deficient cases, when categorized into the mechanism of MMR loss, 15 patients had Lynch syndrome, 98 had MLH1 promoter methylation, 13 had biallelic somatic mutations, and 18 without results available as of yet. Now, looking at the two cohorts and comparing their clinical characteristics, the woman with MMR deficient tumors tended to be older. The stage distributions were similar between the two groups. MMR deficient tumors tended to be higher grade with a propensity for lymphobascular space invasion. MMR deficient tumors were also more likely to have lymph node involvement, and the patients were more likely to have received neoadjuvant and adjuvant therapy. Looking at the unadjusted Kaplan-Meier survival curves, there is no difference in the overall survival between the MMR deficient and MMR intact cohort, but a significantly worse recurrence-free survival in the MMR deficient group after a medium follow-up of 16.8 months. Now, when adjusted for covariates such as age at diagnosis, stage, lymphobascular space invasion, use of adjuvant treatment, and post-upgrade distributions, MMR status was no longer predictive of worse recurrence-free survival. Now, comparing the recurrence patterns, the MMR deficient tumors were more likely to recur in the retroperitoneal lymph nodes, whereas MMR intact tumors were more commonly likely to recur in the distant sites. Now, when analyzing survival outcomes based on their mechanism of mismatch repair loss of the MMR deficient cohort, patients with the MLH1 methylated tumors had the worst recurrence-free survival. On multivariable analysis, after adjustment for age and adjuvant therapy, MLH1 promoter methylation status was still associated with worse recurrence-free survival. The main takeaway points are that, number one, MMR deficiency is associated with older age, higher grade, propensity for lymphobascular space invasion, lymph node involvement, as well as a shorter recurrence-free survival. The mismatch repair deficient endometrial cancers were more likely to recur in the retroperitoneal lymph nodes, and that MLH1 methylated endometrial cancers demonstrated the worst three-year recurrence-free survival. Now, in keeping with our data, there has been a couple of studies similarly showing that MLH1 methylated endometrial cancers tend to have greater tumor volume and worse outcomes. Now, this is not the only study. Now, this Leon-Castillo study in 2020 looking at the PORTEC3 population and response to adjuvant therapy based on the molecular subtype really highlights the fact that MMR deficient endometrial cancers may be less chemosensitive, and we really need to start individualizing treatment based on the molecular subtype. Similarly, although our study is hypothesis generating in nature, there may be some evidence based on our study to consider sampling lymph nodes in early grade MLH1 deficient cases given their worst recurrence-free survival outcomes. Thank you very much, and I'd like to thank all our collaborators that made this study possible, and especially a big thank you to Dr. Sarah Ferguson. Great. Thank you so much, Dr. Kim. What an excellent presentation. Great work. Next up, we're bringing back up Dr. Emily Thompson, and for her second presentation of the morning, Dr. Thompson, please load your slides. Thank you very much. Thank you very much for that introduction. This presentation will focus on variation in clinical practice encountered in this cross-Canada study. I have nothing to disclose. There was wide variation in clinical practice and outcomes for endometrial cancer with high grade histology. None of the five major risk stratification strategies shows high accuracy in predicting the risk of recurrence or nodal metastases in patients with early stage disease. We've already discussed today the role of molecular subtyping and stratifying cancers into four prognostically relevant groups. There is also further evidence of predictive value, which has already been touched upon today. Mismatched repair-deficient cancers have been shown to have improved response to adjuvant radiotherapy, and FDA has recently approved immunotherapy for mismatched repair-deficient cancers, including mismatched repair-deficient endometrial cancer. The recently published PORTEC3 results showed significantly improved outcomes for women with P53 abnormal cancers who received chemotherapy in addition to radiotherapy, regardless of histology. Overlapping survival curves in mismatched repair-deficient cancers in this study show us that these patients do not benefit from the addition of chemotherapy in the setting of advanced adjuvant-treated disease. Poly-mutated cancers in this study had excellent outcomes in both trial arms. Taken together, these findings point to the need for tailored treatment interventions. In this study, we set out to assess current clinical practice across diverse Canadian endometrial cancer cohorts and hope to identify potential impact of molecular subtype-directed care, including projected cost implications. This site depicts participating sites across Canada, and from each of these centers, we received FFPE tissue from either a biopsy or hysterectomy specimen for all cancers diagnosed or treated in 2016. Spoken to the molecular subtyping that was performed in clinical pathologic data gathered included whether the patient was managed by specialists or generalists at the time of surgery. We also collected data regarding surgical treatment and follow-up parameters, as well as hereditary cancer program referrals. 862 cancers are included in this analysis, and they originate from seven major cancer care centers and seven community hospitals. This slide outlines the variation in surgical and post-surgical management for patients across the various participating sites, as depicted on the x-axis. You will recall that site 3 represented community-based care, P53 represented community general hospitals in British Columbia. You'll also remember that sites 4 and 5 had that high predominance of P53 abnormal disease, and so it's not surprising that they undertook more surgical interventions in their patient population. But across the board, we do see marked variation in lymph node and omental sampling, and conspicuously, the extent to which patients received surgical interventions did not necessarily track with the likelihood that patients were followed up in specialist care centers post-operatively. This slide depicts adjuvant treatment differences in mismatch-prepared deficient cancers. As you can see, again, we see wide variation by stage. I'd like to draw your attention to patients in stage 2, 3, and 4 on the right side of the screen. 37% of these patients did not receive radiation therapy as part of their post-surgical treatment. This slide shows adjuvant treatment differences in P53 abnormal disease. As, again, we see wide variation in practice, and I would like to draw your attention to those patients in stage 2, 3, and 4 in whom 22% of cases did not receive chemotherapy, which we now believe to be of value for these patients especially. In terms of HCP, hereditary cancer referral patterns, local mismatch repair testing recognized only 25% of mismatch repair deficient cases in our series. As I mentioned, 11.5% of these did not undergo further evaluation through hypermethylation testing or hereditary cancer program referrals. 10% of all cases were referred to hereditary cancer programs, but of these, over 40% were MMR proficient. Seven patients were identified as having Lynch syndrome, but based on the prevalence of Lynch syndrome and endometrial cancer, nationwide practice in 2016 significantly under-recognized patients with Lynch syndrome in the context of their care for endometrial cancer. You will be aware of prospective trials currently enrolling patients such as PORTEC4A and the Canadian multi-site taper trial. These are assessing the added value of integrating molecular subtyping to risk assessment. We identified 63 patients in our retrospective study who received post-surgical treatment and in whom, based on inclusion criteria to these prospective trials, would have been candidates for observation or vaginal brachytherapy only. De-escalation in these patients would have resulted in a cost reduction of between $350,000 to $400,000 Canadian dollars or between $5,500 and $6,500 Canadian dollars per patient, which includes the cost of molecular classification. In summary, we encountered significant variation across all aspects of endometrial cancer management in our study. PROMIS identified missed therapeutic opportunities. I hope I've managed to convince you that molecular subtyping offers uniform and reproducible classification, which can enable progress towards better outcomes by studying biosimilar tumors in clinical trials, evaluating treatment efficacy. Thanks again to all of our collaborating institutions and to my supervisors and to the Michael Smith Foundation for Health Research. I'll be delighted to take your questions when the time presents. Thank you. Thanks once again, Dr. Thompson. What great work. It's really been my pleasure to moderate this portion of the plenary session. Now we'll hear a five-minute distillation from Dr. Karin Kreuzberg, a radiation oncologist, world-renowned radiation oncologist from Leiden University, Netherlands. After her talk, the second moderator, Dr. Renee Borela, will take over the session. Thank you so much. I'll now turn the screen back to Dr. Kreuzberg. Thank you. Thank you so much for your introduction. I hope you can see on my screen now. I'll discuss these three excellent presentations which have shown that we're now entering the molecular era in endometrial cancer. I have nothing to disclose. Only seven years ago, the Cancer Genome Atlas published their genome-wide analysis of endometrial cancer, showing four molecular subgroups, the Pol E ultramutated group with excellent prognosis, the copy number high driven mostly by P53 mutations with a inferior prognosis, and the MSI hypermutated or mismatch repair deficient subgroup, and the copy number low and the metroid or no specific molecular profile group with both prognosis in between. Dr. Emily Thompson has shown clearly the Vancouver results which are in fact detecting the molecular subgroups published by the Cancer Genome Atlas by their surrogate markers in paraffin-embedded tissues, so to be used in daily clinical practice. Immunohistochemistry can detect it for mismatch repair proteins. P53 abnormal expression can be detected, and DNA sequencing is needed for Pol E excellent mutation, and if none of these are found, the tumors have no specific molecular profile. Both groups in Vancouver and our group have shown that detecting the groups by their molecular surrogate markers to find molecular subgroups has shown a similar prognostic significance as displayed by the Cancer Genome Atlas, and we should be aware that some of the cases, about 3%, have multiple classifying alterations, and as Dr. Thomas already showed, she expressed that the subclonal expression of P53 can harbor a Pol E mutation, and the Pol E mutation is the way this patient's tumor will behave, and not the subclonal P53, so it's essential to do all of the molecular tests to know if it's a multiple classifying alteration or a single molecular subgroup. So Dr. Thompson showed us data for 862 patients, really an excellent effort in a large series, and she found that P53 abnormal and in vitro cancers are found across the range of histotypes, including low-grade and in vitro cancer, and that again shows us that we should look for P53 abnormal cancers because our analysis also confirmed that they behave as P53 abnormal cancer with a worse prognosis, and they are not like their endometrioid counterparts. Subclonal staining was observed in 3.9% of cases, so that also confirms the percentage of double classifiers, so we should look for all of the molecular subgroups, and the molecular subtype was a significantly associated clinical outcome, even in low-stage disease, and I think this is an essential message that we find and use our unexpected P53 staining or mismatch repair deficient staining in all of the cancers, even those of low stage, and that molecular classification will help us to give them the appropriate treatment. Dr. Kim then focused on mismatch repair deficient cancers and analyzed the mechanism of MMR loss with 12% Lynch syndrome within this group, 78% is the vast majority they have mismatch repair deficiencies due to MNH1 promoter hypermethylation. This test has always been done if MNH1 is not expressed. If MSH6 or MSH2 loss is found or isolated PMS2 loss, the patient should be tested for Lynch syndrome, and about half will have Lynch and the other half will be due to bioanalytic somatic mutations, so this is essential for patient management. Her conclusion showed also that mismatch repair deficiency is associated with older age, higher-grade and infested space invasion, and a worse outcome. It is mainly due to the majority of the promoter hypermethylated mismatch repair deficient cases, while those with Lynch syndrome tend to be younger and have a better prognosis. We should be aware of the worst prognosis with the methylation group, and again, it is important to be able to do individualized management based on the molecular group, as these seem to behave differently, and Dr. Kim showed the lower recurrence with survival of the methylated subgroup in comparison to those with Lynch syndrome or intact mismatch repair deficient proteins. Then Dr. Thompson also showed in the same group of patients that currently surgical staging, adjuvant therapy and catheterization are very inconsistent, widely different between the different practices, and these variations are irrespective of the stage or group, and that the introduction of molecular subgroup-based treatment would help us to get better and more effective treatment for those in need of adjuvant treatment, but also allow de-escalation which would greatly reduce cost, and it's important to know this because in discussions you will see that polyimutation is a costly procedure to do in daily practice, and that immunohistochemistry, of course, not that expensive, but DNA sequencing is, but you can save many more costs by de-escalating treatment when possible than you would spend in diagnosing the four molecular subgroups. So the classification, and I fully agree with Dr. Thompson that this can provide consistency in care and direct biologically informed management. In the PORTEX-3 trial of only high-risk endometrial cancers with 30% stage 3, 30% grade 3, so really a group of high-risk cancers, there was a relatively high percentage of P53, 22%, and about one-third were mismatch repair deficient, and 30% POE, and again across these high-risk cancers, a wide difference in prognosis was found with poor prognosis for the P53 and excellent prognosis for POE even in this high-risk subgroup. So it is essential not to consider these uniformly high-risk, but to do molecular group-based diagnostics and treatment, especially because we found important differences between the treatment arms in the four molecular subgroups with, in both arms, with or without chemo, the same excellent relapsing survival for POE mutant, which gets you wondering if they need any adjuvant treatment at all. The significant and clinically relevant improvement of relapsing survival with addition of chemotherapy to the radiation for P53, but no difference for mismatch repair deficient, and it was unexpected. We all know, and Dr. Thompson already told us, that these might be more responsive to checkpoint inhibition, so in advanced stages this could be the way forward, and those with no specific molecular profile had, in this smaller group, a non-specific, not significant, but a trend towards better survival with addition of chemotherapy, similar to the complete trial outcome. But for this group, this is highly relevant, that chemo might not be working. So molecular typing is the way forward, and it's here to stay. We will not see treatments in the coming ages without, and the WHO has already paved the way for this. Immunohistochemistry and POE testing should be incorporated in any clinic in routine diagnostics and treatment decisions. We could use these factors, integrating them with other strong factors, because factors like substantial lymphoscopic base invasion, beta-catenin mutation, LNKAM and HART2 are also of significance, and integrated profiles seem more effective. So more research is needed, and that will more others come out of the Project 4 trial, but we should target specific activation for more effective treatments, and we need clinical trials. Examples are the Canadian TAPER trial, looking at reduction of adjuvant treatment when possible, based on p53 wild type, beta-catenin wild type, and no extensive LVSI, and for POE recommend no treatment. And currently, the plan is to add stage 2 and 3 POE mutated cancers to this in the scope of the International Rainbow Trials Program, where Canada will lead to POE trial without adjuvant treatment, but every single molecular group will have their own informed trial question for p53. PARP inhibition for added value, for mismatch repair deficient, if we would add checkpoint inhibition, and for those without a specific molecular profile, that we could substitute chemo for hormonal treatment, because these are uniformly hormone receptor positive. So I think we are now on the verge of doing molecular-based treatment for all of our patients, and they will greatly benefit. Thank you so much for your attention. Thank you, Dr. Kretberg. Hello, everyone. I'm Dr. René Pareja from Instituto Nacional de Cancerología in Bogotá and Clínicas de Orgán en Medellín, and I'll be moderating the second half of this plenary session. I'd like to remind you that at the end of the presentations, we will have a live question and answer segment. Please submit your questions with a Q&A feature at the bottom of your screen, and we will do our best to answer them. The next abstract presentation is from Maria Cusimano. Dr. Cusimano, please present your abstract now. Thank you. Thank you for the opportunity to present today. I will be sharing our study on sentinel lymph node biopsy versus lymphadenectomy for intermediate and high-grade endometrial cancer, staging the Centaur Prospective Multicenter Cohort Study, and we have no conflicts to disclose. Sentinel lymph node biopsy has gained acceptance in low-grade endometrial cancer, but its role in high-grade endometrial cancer is still debated. Prospective studies published to date have enrolled few patients with high-grade histology, and so the diagnostic accuracy of sentinel lymph node biopsy in this specific patient population has not been definitively established. To address this, we performed the Centaur Prospective Multicenter Cohort Study at three academic hospitals and designated cancer centers in Toronto, Canada, and the Centaur study was specifically powered to evaluate the diagnostic accuracy of sentinel lymph node biopsy using endocyanin-green fluorescence imaging in patients with intermediate and high-grade histology. We enrolled clinical stage 1, grade 2 endometrioid, and high-grade endometrial cancer patients undergoing laparoscopic or robotic primary hysterectomy with an intent to complete full staging. We excluded patients with grade 1 endometrioid histology, patients with evidence of advanced or metastatic disease on preoperative imaging or at the time of an initial intraoperative survey, and those with contraindications to either lymphadenectomy or ICG. All patients underwent a two-step algorithm for sentinel lymph node biopsy. In the first step of the algorithm, patients underwent cervical injection of ICG followed by retroperitoneal dissection of mapped sentinel lymph nodes, and then in the second step of the algorithm, patients underwent side-specific lymphadenectomy for non-mapped hemipelvises if that was applicable to them. All sentinel lymph nodes were handled using a standardized ultrastaging protocol at all centers, which involved use of both H&E and immunohistochemistry. Patients then underwent the reference standard. For patients with grade 2 histology, this entailed bilateral pelvic lymphadenectomy, and for patients with high-grade histology, this entailed both bilateral pelvic and bilateral periodic lymphadenectomy. Our primary endpoint was sensitivity of the sentinel lymph node biopsy algorithm, which was defined as the proportion of node-positive patients correctly identified by the algorithm. For these calculations, patients were deemed true positives if metastases were correctly identified either on the sentinel lymph node specimens from mapped hemipelvises or on the side-specific lymphadenectomy specimens from non-mapped hemipelvises. We chose this outcome because we felt it was most clinically relevant to individual patients and decisions on whether or not to administer adjuvant therapy. Sample size was determined based on our primary endpoint, which was sensitivity of the algorithm, and we used a Fleming two-stage design to test the null hypothesis that the sensitivity was 80 against a one-sided alternative that the sensitivity was 93. We required a total of 46 node-positive patients recruited in two stages to test this hypothesis. In the first stage, we enrolled 25 node-positive patients. The study would stop early for futility if 20 or fewer patients were correctly identified by the algorithm or stop early for efficacy if 24 or more were accurately identified by the algorithm, and this interim endpoint was actually met in June 2019. From July 2015 to June 2019, we enrolled 156 patients with a median age of 66 and a median BMI of 28, and 126 patients in the cohort, or over 80 percent of the cohort, had high-grade histology. Twenty-seven patients had nodal metastases, so 17 percent of the cohort. Twenty-four of these were in patients with high-grade histology, and only three were in patients with grade 2 endometrioid histology. This total of 27 includes the 25 patients that triggered initial stoppage of the study in June 2019, as well as an additional two patients who had undergone surgery prior to that point and whose pathology reports became available only after our interim analyses were complete. As I mentioned, all patients underwent cervical injection of ICG. Detection rates of sentinel lymph nodes were 97 percent per patient, 88 percent per hemipelvis, and 78 percent bilaterally, and surgeons removed a median of three sentinel lymph nodes per patient. Patients then underwent the reference standard. All patients received pelvic lymphadenectomy, and 80 percent of high-grade patients additionally underwent periodic lymphadenectomy. The median number of pelvic lymph nodes retrieved was 16, and the median number of periodic lymph nodes retrieved was 5, again both indicating adequate technique for lymphadenectomy. For our primary analyses for the diagnostic accuracy of the sentinel lymph node biopsy algorithm, all 156 patients were included in that analysis. 26 of 27 node-positive patients were correctly identified by the sentinel lymph node biopsy algorithm, resulting in a sensitivity of 96 percent, a false negative rate of 4 percent, and a negative predictive value of 99 percent. Only one patient in our cohort, or 0.6 percent of the cohort, would have been misclassified if they received sentinel lymph node biopsy alone. In terms of the positive cases, so of the 27 node-positive patients identified, 14, or 52 percent, had metastases in the sentinel lymph node specimen only, and all 14 of these patients had either micrometastases or ITCs. Two patients had a single positive sentinel lymph node found outside of traditional pelvic lymphadenectomy boundaries, and five patients required immunohistochemistry for diagnosis. So a total of seven patients, or a total of about 26 percent of all node-positive patients, may not have been identified on pelvic or periortic lymphadenectomy alone. To conclude, sentinel lymph node biopsy has excellent performance characteristics in patients with intermediate and high-grade endometrial cancer. In our cohort, the false negative rate of a sentinel lymph node biopsy algorithm was only 4 percent. What our study really adds is looking predominantly at the high-grade patient population, over 80 percent of our cohort, and 89 percent of node-positive patients identified had high-grade histology. So together with existing literature, our findings suggest that sentinel lymph node biopsy is a viable replacement for lymphadenectomy in terms of identifying lymph node metastases, and this is the case for both low- and high-grade endometrial cancer. I'd particularly like to acknowledge Dr. Sarah Ferguson, who is the principal investigator on the SANTOR trial, for really taking us through from start to finish, as well as all the surgeons, pathologists, and research staff who made the study possible. Thank you. Thank you, Dr. Cusimano. That was a very interesting study. Congratulations on your work. Next, we will hear from Dr. Salvatore Guagliaretti. Salvatore, please begin your presentation. Okay, good morning to everyone. It's really a big pleasure for me to be here and an honor to have the possibility to share our result from this multi-centric randomized trial that we designed in 2015 to evaluate the role of uterine manipulator on laparoscopic or robotic hysterectomy for the treatment of low-risk end-of-life cancer patients. That was named the Romani trial. I have no conflict of interest to disclose. And in 2015, when we designed, although the role of intrauterine manipulator have already been widely discussed in terms of pathological and oncological outcome, the available literature was heterogeneous in terms of study design, types of manipulator used and evaluated during the study, and aim. There was only one randomized study at the time with heterogeneity in terms of preoperative histotype, so patient population that can be considered delimitation. So we decided to design this multi-centric randomized trial involving five Italian tertiary care centers. My hospital, the Fondazione Politinico Gemelli of Rome, was the coordinating center. And the inclusion criteria was presumed. Low-risk end-of-life cancer patients had preoperative workup. This was the target of patients. So the primary objective was to assess the effect of uterine manipulator in terms of lymphomodular space invasion, lymphovascular space invasion and peritoneal cytology. And the secondary objective was to evaluate the preoperative and oncological outcome. During the study period from October 2015 to December 2017, 227 patients were evaluated for eligibility. Finally, 154 patients were randomly allocated in two different study groups, the group A in which we used the uterine manipulator and the group B in which we didn't use the uterine manipulator. Both patient populations, both groups were very balanced. I'm sorry, but it's really difficult to present all the results. We were taking into account the preoperative variables. We identified two statistically significant differences in both groups of patients. The first was regarding the operative time. To better understand, we conducted the sub-analysis dividing patient on the basis of the approach laparoscopic or robotic. And the difference was related to the laparoscopic procedure, probably because the high experience of surgeon involved in this study determines the presence of a uterine manipulator or not something not so useful. On the other hand, we had another statistically significant difference between the two groups regarding the estimated blood loss. But even here, the increased AVL in the group A, so the uterine manipulator group, was due to a vascular complication during lymph nodal assessment, so nothing related to the uterine manipulator. For what concerns the pathological finding, no difference between both study groups. In terms of lymphovascular space invasion, in terms of pattern of positive lymphovascular space invasion, in terms of cytological finding, we performed in all patients pre-hysterectomy cytology before to manipulate the uterus and the post-hysterectomy cytology after the vaginal cap closure. We never had a positive cytology in our patient population. For what concerns the oncological outcome, finally, the intrauterine manipulator did not afflict the DFS at the univariable and multivariable analysis. Differently, the combination between clinical stage and histology afflicted the DFS, but this is nothing new. The uterine manipulator did not afflict the DFS. This is reinforced because median overall survival and median disease-free survival with a median follow-up time of 38 months was no difference, even if we don't reach the DFS. For what concerns the conclusion, we can condensate the conclusion in two different aspects. The surgical one, the use of intrauterine manipulator is not related to an increased risk of intrapartic complication. In the laparoscopic procedure, the manipulator's installation time represents a time-consuming event without any significant surgical advantage when the surgeon is high-experienced. This is really important to underline because uterine manipulator is something that really assists the surgeon in exposing the anatomy, both superficial and retroperitoneal. This is partially confirmed by the fact that in the robotic procedure, the FORTAR probably effectively compensates the absence of the uterine manipulator in exposing the surgical structures. In fact, we didn't record any difference statistically significant, obviously, between both operative time in both groups. For what concerns the oncological aspect, the use of intrauterine manipulator has no impact on both the cytology and LWSI status. There is no association between the uterine manipulator and the risk of recurrence and it doesn't afflict influenced DFS and OS. Uterine manipulator is something useful in surgery, but it's very important to perform a good assessment of preparatory risk factor in parallel with the correct evaluation of the patient to avoid any surgical and oncological artifact due to the presence of a usage of a uterine manipulator. That should be considered a weaver and important element of a personalized surgical treatment. With this, I want to thank you for your attention. I want to thank all my friends and colleagues for driving together this paper. Thank you. Thank you, Dr. Arletty, for sharing your work with us. Now, we will move on on the final presenter of this session, Dr. Morkos Nakhla, and then we will open up for questions from our audience. Please begin, Dr. Nakhla. Hello, everyone, and thank you for being here. My name is Morkos and I'm a second year medical student at UCLA. I'm honored to be here with you virtually. Today, I will present our project on endometrial cancer. Specifically, we find that frailty is independently associated with worse outcomes and increased resource use following procedures to treat endometrial cancer. I have no disclosures. What is frailty? Frailty is a syndrome of decreased resilience to stress. It involves an increased burden of comorbidities, decreased physiological reserve, and an increased vulnerability even among same-age individuals. Therefore, frailty has been associated with worse clinical outcomes following many procedures. Ways to assess frailty preoperatively may inform decision-making and guide perioperative care. The main objective of our study was to assess the impact of frailty on surgical outcomes and patients specifically undergoing operations for endometrial cancer. To do this, we used a large administrative database. We used the years 2005 to 2017 from the National Inpatient Sample. This represents the largest collection of administrative healthcare data in the United States and is a representative sampling of all inpatient hospital encounters. To define frailty, we used the Johns Hopkins Adjusted Clinical Groups Frailty Defining Diagnosis Indicator. This is a multidimensional instrument used to assess frailty and has been validated for research using administrative data. This frailty indicator uses 10 clusters of diagnoses that define someone as frail, including malnutrition, impaired vision, dementia, and poverty. Using our database, we found about 380,000 patients who were diagnosed with endometrial cancer. After making some exclusions, including those under 18 years of age and those missing key pieces of information, and only keeping patients who underwent one of these procedures during their hospital stay, we divided our cohort into those considered frail and those who were not frail. Frail patients made up 2.9 percent of our cohort. We then analyzed these patients according to frail status. Those who were frail were on average older and had a greater burden of comorbidities according to the Elixhauser Index. The Elixhauser Comorbidity Index is a thoroughly validated method of assessing comorbidities using administrative data. Looking more specifically, we find that frail patients were more likely to have certain comorbidities, including congestive heart failure, diabetes, chronic lung disease, and obstructive sleep apnea. Frail patients, however, were less likely to be obese. Frail patients were also more likely to undergo an extended procedure, meaning more than a standard staging procedure. For example, patients who received a bowel resection. Frail patients were also less likely to receive minimally invasive procedures. In terms of outcomes, when adjusting for patient and hospital characteristics via a multivariate regression analysis, frail patients had increased odds of developing several comorbidities, including neurologic, respiratory, infectious, and renal complications. Frail patients were also significantly more likely to have increased odds of mortality, as well as non-home discharge. Finally, being frail was associated with a nearly three-day increase in length of hospital stay and likely related a nearly $8,000 increase per patient in hospitalization costs. After adjustment for relevant variables, including age, we also find that frailty predicts a significantly higher rate of mortality in patients with older age, the cutoff seeming to be right under 50 years of age. We do also find, however, that the rate of mortality decreases significantly for frail patients throughout our study period, which is about 12 years. There are limitations to our project. These include inherent limitations of coding and undercoding, a lack of stage and histopathology data, and a lack of readmission and outpatient follow-up data inherent to the administrative database and its characteristics. We do, however, use a large national data set, which represents an estimate of all inpatient hospital encounters in the U.S., and we analyzed 12 years of this data. In conclusion, we find that frailty is independently associated with worse surgical outcomes in women with endometrial cancer. We also find increased mortality is associated with frailty in this cohort. Finally, we find reduction in mortality for frail patients over the study period. In future directions, we hope that further efforts to mitigate the impact of frailty are explored. We also hope to understand the biological changes associated with frailty. A next step could be a prospective study of patients with endometrial cancer who are considered frail, and our ultimate goal could be the development of a shared decision-making model for frail patients with endometrial cancer. I would like to thank you all for your time. Thank you very much, Dr. Nakla. Now, we'll open up for questions from the audience for five minutes. The viewers have been submitting questions throughout the presentations, and we have selected a few. The first question is directed to Dr. Emily Thompson. What is the potential role of immunohistochemistry in molecular profiling in low-resource settings? Thank you very much for that question. I guess the first part of that question is identifying cancers first in which knowing molecular subtype would add something for the patient. For patients who have early stage and low-grade disease, molecular subtyping may not be of benefit for those patients. For those with intermediate to high-risk cases or disease, immunohistochemistry can be done for P53 and two-panel mismatch repair gene immunohistochemistry can be done, so consisting of MSH6 and PMS2. That should help you identify those intermediate and unfavorable risk cancers. We acknowledge that polymutation testing won't be widely available for some time, but there will be work that's done to try and identify the most intelligent way to apply what testing we have available. Thank you, Dr. Thompson. We have a question for Dr. Cusimano about the value of isolated tumor cells in the future because you had a group with this finding. Sorry? What are your thoughts about isolated tumor cells and the role that those cells can be in the future in the management of sentinel node detection in endometrial cancer? Our study definitely wasn't designed to answer that question. We did consider node-positive to be macrometastases, micrometastases, or ITCs. The patients who were node-positive in our cohort, some of them did have isolated tumor cells, but in terms of the long-term prognosis and how that should factor into management, that certainly wasn't something our study was designed to answer. I think the role that ITCs play, particularly in high-grade endometrial cancer, is not clear. I think it is important that we continue to identify those patients, at least until we have a better understanding of their prognosis. I think identification is still important. I think we've shown with our study that those patients are appropriately identified with sentinel biopsy. Thank you, Dr. Cusimano. I have a question for Dr. Nakla about what do you think the related mortality due to frailty has decreased over time? Great question. I would think this might be perhaps due to recent alternative management strategies, like using adjuvant chemotherapy prior to surgery, or an optimization of comorbidities that led to this decrease in frailty over the study period. Okay, thank you very much. Unfortunately, due to schedule issues, we don't have more time for answering questions from audience. We need to end the session. I'd like to thank all the presenters for sharing their excellent work with us. If there are any questions that were not answered here, please feel free to contact the speakers through the IGCS link here in the meeting portal.

endometrial cancer

molecular subtyping

sentinel lymph node biopsy

frailty

surgical outcomes

molecular classification

prognostically relevant risk groups

diagnostic accuracy

intrauterine manipulator

laparoscopic hysterectomy