Fullständig transkription av Beyond the Binary: Unpacking Molecular Diversity in PMMR Endometrial Cancer

outside of the classic biomarker for PARP inhibitor, which is, of course, homologous recombination deficiency, which is a whole other conversation in endometrial cancer. 00:02:39.120 --> 00:02:43.280 So the design of DuoE, we enrolled 718 patients. 00:02:43.439 --> 00:02:53.280 They're randomized one-to-one-to-one, double-blind, placebo-controlled, global, with newly diagnosed advanced or recurrent endometrial cancer. 00:02:53.360 --> 00:02:55.759 It included carcinosarcomas notably. 00:02:55.919 --> 00:03:00.159 And we did central mismatch repair by IHC before randomization. 00:03:00.240 --> 00:03:15.680 So it was a key stratification factor and randomized patients to carboplatin, paclataxol alone, carboplatin, paclataxal, drvalumab, drvalumab maintenance, and then carboplatin, paclataxil, draval, draval magnets. 00:03:15.840 --> 00:03:35.680 And both experimental arms, the derva for dervalumab, and dervalumab olaprib, met the dual primary PFSN point with hazard ratios of 0.71 for derva alone and 0.55 for derva olaprib, both versus the control of chemo alone. 00:03:36.000 --> 00:03:43.520 By mismatch repair, of course, derva's largest effect was in DMMR with a hazard ratio of 0.42, and that's not surprising. 00:03:43.680 --> 00:03:50.000 But in PMMR, Derva alone gave a hazard ratio of 0.77, which is, I think, clinically relevant. 00:03:50.159 --> 00:04:00.639 But adding Olapra maintenance here moved that hazard ratio to 0.57 with a confidence interval that was pretty tight, 0.44 to 0.73. 00:04:00.800 --> 00:04:04.800 So we zeroed in on this PMMR subgroup. 00:04:04.960 --> 00:04:08.400 It's the large majority of patients enrolled to the clinical trial. 00:04:08.560 --> 00:04:14.800 And it's the group where Olapra of addition looked like it contributed the most incremental value. 00:04:15.120 --> 00:04:20.079 DMMR didn't really benefit at all because they had high immune checkpoint effectiveness. 00:04:20.319 --> 00:04:27.040 So we wanted to understand more about the molecular kind of texture of this PMMR group. 00:04:27.120 --> 00:04:30.800 Because I think we're going to talk about PMMR isn't one thing. 00:04:31.040 --> 00:04:44.639 So who were the patients, who are their, what were their tumors, and whether benefit was concentrated in one subset of this or broadly across all molecular subsets of PMMR. 00:04:44.959 --> 00:04:54.079 And so the biomarkers that were selected, we had to pick some, we couldn't pick everything, were based on the promise groups, the promise-defined groups. 00:04:54.160 --> 00:05:00.319 So poly, TP53 are those that are increasingly used in practice and carry a lot of prognostic weight. 00:05:00.480 --> 00:05:01.680 So we anchored on those. 00:05:01.839 --> 00:05:07.040 And then PDL1, of course, is a predictive marker for anti-PDL1 therapies. 00:05:07.199 --> 00:05:13.439 We looked at homogenous recombination repair mutations because HRD is not a validated marker yet. 00:05:13.519 --> 00:05:17.680 And BRACA, because those

are associated with, of course, PARP inhibitors. 00:05:17.759 --> 00:05:21.279 And then there were some others sort of sprinkled in, but those were the main ones. 00:05:21.680 --> 00:05:22.399 That's great. 00:05:22.560 --> 00:05:38.639 You know, as your manuscript title implies, you observed a great deal of biological heterogeneity in this broad class of endometrial cancer patients with PMMR disease and with several potentially actionable mutations in your study cohort. 00:05:38.800 --> 00:05:48.800 Was that surprising to you that there was such a high rate of biomarker expression in this PMMR cohort and that more than half expressed two or more biomarkers? 00:05:49.279 --> 00:05:55.199 Well, I think that we had hints of this from some of the readouts that we were seeing from Dr. 00:05:55.360 --> 00:05:58.160 Eskander's study, from Ruby. 00:05:58.240 --> 00:06:13.040 We were starting to see hints that of what we, I think we've all known that PMMR is a heterogeneous group at a more kind of probably a more varied set of tumors than just this dichotomization of DMMR, PMMR. 00:06:13.199 --> 00:06:40.160 But I do think it was somewhat surprising and a little humbling to see just how heterogeneous this group is and the overlap of biomarkers, I think it's going to be a common thing that we talk about in a lot of our tumors because it's going to challenge us to really understand the predictive value of each of these and in the context of the presence of one another to best individualized therapies for our patients moving forward. 00:06:40.319 --> 00:06:44.720 So as you alluded to, in now, this was a subset analysis. 00:06:44.800 --> 00:06:52.000 So we had 84% of the biomarker evaluable population was positive for at least one of the markers we were looking at. 00:06:52.160 --> 00:06:53.920 So it's pretty common to find one. 00:06:54.160 --> 00:06:59.600 53% of our biomarker evaluable population had two or more. 00:07:00.000 --> 00:07:02.959 So the majority of patients had two or more biomarkers. 00:07:03.040 --> 00:07:16.399 And the most common were PDL1, which I'm not sure we care about here, and TP53, which we may start to carry about a lot at 67 and 59%, respectively, and 44% had both of those. 00:07:16.800 --> 00:07:28.240 So I think the framing here is that you reframe PMMR as a more molecularly diverse bucket rather than a single entity. 00:07:28.480 --> 00:07:33.279 And the overlap, I think, is going to be the story here in terms of how we move forward. 00:07:33.519 --> 00:07:42.480 I'll point out the thing that's missing from this analysis, because we all are going to kind of throw rocks at our glass houses, is high two is not part of the analysis. 00:07:42.639 --> 00:07:43.920 That would be a logical thing. 00:07:44.000 --> 00:07:59.040 I wish we had it so that we could look at the likely TP53 per two positive group and then see how Olaparib worked there, given the rise of TDXC, and you're going to potentially need to be making decisions for that group of patients. 00:07:59.199 --> 00:08:06.560 So there are missing things from this analysis that we can all always look back on things and wish we'd done them a little differently. 00:08:06.879 --> 00:08:12.160 Yeah, balancing scientific idealism with, you know, pragmatic, getting the biomarker work done. 00:08:12.639 --> 00:08:13.199 I get it. 00:08:13.439 --> 00:08:19.040 But the contributions are really meaningful here, I think, and very thought-provoking about what do we do next. 00:08:19.279 --> 00:08:20.639 And so I'll turn now to Dr. 00:08:20.800 --> 00:08:21.439 Eskander. 00:08:21.680 --> 00:08:30.720 You you authored a very thoughtful and elegant commentary to this work and you know discussed a lot about tumoral biomarker validation. 00:08:30.879 --> 00:08:39.360 So, which tumoral biomarkers are validated currently in endometrial cancer and how should we interpret the current biomarker analysis we're discussing on the podcast? 00:08:39.600 --> 00:08:45.279 Where do we go from here with this provocative data and potential future validation of exploratory biomarkers? 00:08:45.759 --> 00:08:48.639 Thank you for the question and congratulations again to Dr. 00:08:48.720 --> 00:09:00.080 Moore and the co-authors for this work because it is very important for us to start to really think about the relevance of these molecular characteristics within the mismatch repair proficient population. 00:09:00.320 --> 00:09:04.320 I will say that this answer might be debated depending on who you ask. 00:09:04.480 --> 00:09:18.080 Although the way I interpret evaluation of these biomarkers, and for clarity's purposes, some can be prognostic for clinical outcomes, some are known to be predictive of response to a therapeutic intervention. 00:09:18.159 --> 00:09:24.320 And then there's going to be a group of biomarkers that may be both prognostic and predictive in terms of clinical outcomes. 00:09:24.399 --> 00:09:29.200 And I think it's important to keep each of those buckets clear in our minds as we have these discussions. 00:09:29.360 --> 00:09:39.519 So currently, biomarkers that we use to inform therapy selection that it's been associated to be predictive of response really hinge on the tumor mutational burden status. 00:09:39.679 --> 00:09:45.120 We know that we have FDA disease agnostic approvals based on tumor mutational burden status high. 00:09:45.440 --> 00:09:58.960 We know that we have also biomarker mismatch repair deficiency, which you can think of as linked to tumor mutational burden, that is also associated with an FDA-approved indication for immune checkpoint inhibition across solid tumors. 00:09:59.039 --> 00:10:04.480 So the mismatch repair deficiency status, the TMB high status, and the MSI high status. 00:10:04.639 --> 00:10:25.440 And one of the things that I will also tie into this is what you'll see as we think about biomarkers predictive of response to a therapeutic intervention, is we see a lot of consistency amongst trials

that are done in a certain disease setting, but even amongst disease agnostic manner, meaning different disease settings, different solid tumors with a therapeutic intervention and a biomarker. 00:10:25.600 --> 00:10:42.320 And so I will add to that also the emergence of HER2, which has existed in the GYN space and the breast, of course, for a period of time, but with the destiny pan tumor O2, now we have again a disease agnostic approval for a therapeutic drug with a HER23 in the NCCN compendium. 00:10:42.480 --> 00:10:44.240 It expands it to 2 and 3 plus. 00:10:44.559 --> 00:10:55.279 But in my mind, those are biomarkers for which we have a therapeutic intervention tied to the presence of that actionable biomarker, which predicts a clinical response. 00:10:55.440 --> 00:10:58.559 And some of these, of course, can also be prognostic for outcome. 00:10:58.799 --> 00:11:04.399 So those are where I believe to be the most mature biomarkers predictive of response right now. 00:11:04.879 --> 00:11:06.159 I really appreciate that. 00:11:06.399 --> 00:11:13.679 Going into our, you know, our real-world clinics here and taking this data forward, I'm curious, and we'll start with you, Dr. 00:11:13.759 --> 00:11:21.200 Eskander, how do you approach molecular testing for your endometrial cancer patients, which biomarkers are routinely tested? 00:11:21.440 --> 00:11:27.840 At what point in the care pathway is there a sequence that you prefer or that is recommended by guidelines? 00:11:28.480 --> 00:11:32.720 It's an important question because it affects all of us in our practices. 00:11:33.039 --> 00:11:42.879 We employed a strategy that would help minimize the risk that we would miss a biomarker that could be important in the care of this patient's treatment paradigm. 00:11:43.039 --> 00:11:53.600 So in our institution, every patient gets mismatropare immunohistochemistry, not because of the mismatropare deficiency implications for therapy, but because of identification of Lynch syndrome. 00:11:53.679 --> 00:11:55.120 And we need to emphasize that. 00:11:55.440 --> 00:12:04.960 We should also mention that the NCCN guidelines for germline genetic testing for endometrial cancer were updated recently, and that's very important because again have implications for patients and family members. 00:12:05.200 --> 00:12:09.840 And we also do ER and PR expression levels done by IHC. 00:12:09.919 --> 00:12:17.279 We do HER to IHC on essentially all non-endometrioid histologies, irrespective of stage. 00:12:17.360 --> 00:12:21.679 And then we do it on advanced stage endometrioid patients and high grade lesions. 00:12:22.000 --> 00:12:30.159 And in addition to that, in patients who have advanced stage and/or any recurrent disease, we also send NGS testing on those patients. 00:12:30.240 --> 00:12:38.080 So independent of the IHC that's being done locally at our institution, we send that tissue out for NGS testing as well. 00:12:38.159 --> 00:12:47.679 So we have that entire data set available to us as we think about treating that patient, both at that setting, and then of course, if they were to develop disease recurrent. 00:12:48.000 --> 00:12:48.639 Yeah, thank you, Dr. 00:12:48.799 --> 00:12:49.279 Eskander. 00:12:49.440 --> 00:12:52.399 I think that's a very, very pragmatic and reasonable approach. 00:12:52.639 --> 00:12:52.879 Dr. 00:12:53.120 --> 00:12:56.639 Moore, did you do something similar at the University of Nebraska? 00:12:56.960 --> 00:13:00.480 Yeah, I think that's I think increasingly a very uh adopted approach. 00:13:00.639 --> 00:13:02.320 We do something very similar. 00:13:02.480 --> 00:13:10.320 And I was very glad that Ramez uh mentioned the new update to the recommendations for genetic testing for endometrial cancer. 00:13:10.480 --> 00:13:18.799 It's probably something we need to talk about as a society more broadly and the implementation of those recommendations. 00:13:18.879 --> 00:13:21.600 Um kind of came out of the blue a little bit for all of us. 00:13:21.759 --> 00:13:24.159 I was like, oh, why that's a good thing. 00:13:24.480 --> 00:13:27.600 But now I gotta like catch everybody up patience. 00:13:27.840 --> 00:13:28.480 That's a lot of patient. 00:13:29.600 --> 00:13:32.000 Yeah, so I think that was a good thing just to bring up. 00:13:32.159 --> 00:13:36.159 Maybe we should say TBD on discussion of that moving forward. 00:13:36.320 --> 00:13:39.919 But I 100% agree with the remainder of that discussion. 00:13:40.080 --> 00:13:54.879 And uh it's incredibly important to develop the sort of pathways within your institutions because everybody's institution looks a little bit different, that allows you to get all of those through whatever mechanisms uh make the most sense for your for your particular site. 00:13:55.360 --> 00:13:56.000 Absolutely. 00:13:56.080 --> 00:13:59.200 Yes, I agree with the new universal testing guideline. 00:13:59.279 --> 00:14:05.279 Uh, it's really going to require a lot of thought and mobilization and commentary about how we move forward with that. 00:14:05.519 --> 00:14:10.960 But it's clear that universal molecular testing is critically important in these patients. 00:14:11.120 --> 00:14:20.240 I'm I'm curious how how you handle poly testing at your centers and at what phase of care you consider doing that. 00:14:20.720 --> 00:14:25.200 I mean, we are not doing poly testing in early stage patients. 00:14:25.440 --> 00:14:32.159 In Oklahoma, where I was previously, we had rainbow open, and so there was testing as a part of that clinical trial. 00:14:32.480 --> 00:14:40.080 But because it's NGS, it's just not something that we're doing without some prompt, you know, currently. 00:14:40.240 --> 00:14:49.360 So it's really something we're seeing when we do NGS in the same setting that Ram is alluded to in this sort of advanced recurrent setting, is where we discover poly. 00:14:49.759 --> 00:14:57.919 Not saying that's ideal, because you would want the

opportunity to de-escalate, maybe a dedifferentiated tumor if it were poly. 00:14:58.159 --> 00:15:02.080 And that might be a place where we would look, but we don't have that trial open. 00:15:02.159 --> 00:15:05.279 And so we are not de-escalating off clinical trial. 00:15:05.519 --> 00:15:12.080 So it's just not something that we're driven to look for right now because we wouldn't act on it in an early stage situation. 00:15:12.879 --> 00:15:14.320 Yeah, I agree completely. 00:15:14.399 --> 00:15:16.639 And I can't emphasize more the comment Dr. 00:15:16.799 --> 00:15:27.919 Moore just made, which is I would never de-escalate a patient who would be indicated to receive systemic therapy based on data results for a pole E mutation off of a clinical trial. 00:15:28.080 --> 00:15:36.559 Because the reality is the data regarding prognostic impact of Pole E is based on retrospectively collected data with patients receiving therapeutic intervention. 00:15:36.639 --> 00:15:46.720 So I think it's really important that those patients be, you know, enrolled on study so that we can definitively answer this question and safely de-escalate in the future if those results end up panning out. 00:15:47.679 --> 00:15:48.159 Here, here. 00:15:48.399 --> 00:15:49.200 Well said. 00:15:49.440 --> 00:15:53.519 Let's pivot a little bit here to talk a bit more about the trial results. 00:15:53.679 --> 00:15:54.080 So, Dr. 00:15:54.240 --> 00:16:16.080 Moore, the original manuscript proposed carbopantlataxilon dervalumab followed by dervalumab and elabrib as a potential new option for patients with PMMR disease, citing uh, you know, pre-planned exploratory uh PMMR subgroup in which adding Elabrib further improved PFS beyond dervalumab maintenance alone for all biomarker subgroups studied. 00:16:16.399 --> 00:16:36.240 However, one of the commentaries or criticisms of that potential approach is that the trial primary endpoint of PFS was only powered to compare each experimental arm to the chemotherapy-only control arm and not to perform a statistical comparison between the two experimental arms, the dervalumab alone versus dervalumab plus elaparib uh arms. 00:16:36.559 --> 00:16:41.679 So, can you comment on what your thoughts on this in terms of practical application? 00:16:41.759 --> 00:16:48.399 And is this a design limitation that it should affect our interpretation of the apparent incremental benefit of a laparib? 00:16:48.960 --> 00:16:53.759 That is the key criticism that we kind of just have to concede. 00:16:53.840 --> 00:16:54.720 You're absolutely right. 00:16:54.799 --> 00:17:01.600 There was no formal statistical comparison between the Durva arm and the Dervo Olaparib arm. 00:17:01.840 --> 00:17:10.559 So the apparent incremental benefit of a laparib is a crossarm descriptive observation, not a tested hypothesis. 00:17:10.720 --> 00:17:22.559 And on top of that, every subgroup analysis in the paper is a post hoc exploratory with no alpha allocation, no type

one error control, and really just point estimates and not p-values. 00:17:22.640 --> 00:17:34.880 So a lot of everything that was in this paper where we say, oh, the apparent benefit extends across all of these subgroups, all post hoc and the comparison of the two arms is descriptive. 00:17:34.960 --> 00:17:39.440 So that is a key criticism that I can't, you know, explain away. 00:17:39.680 --> 00:17:49.599 So what I would say, I think what raises confidence that the Olaparib signal is real rather than just noise, uh, is the consistency. 00:17:49.839 --> 00:18:04.720 So across the biomarker known uh and the histology subgroups, every hazard ratio point estimate favored the treatment arm over control, and derva plus Olaparib was numerically better than derva alone across nearly all of them. 00:18:04.880 --> 00:18:14.160 A signal that holds direction across many independent cuts is harder to dismiss than a lone subgroup outlier, you know, to quote Rob Coleman. 00:18:14.480 --> 00:18:23.200 So I think it's biologically coherence, which is the part plus immunotherapy rationale we opened with. 00:18:23.599 --> 00:18:31.680 So the combination is an approved regimen based on, at least in Europe, based on the trial's intention to treat results. 00:18:31.839 --> 00:18:43.759 And I think where reasonable people can differ is exactly the incremental Olaparib and PMMR question, because that piece rests on a pre-planned but statistically untested comparison. 00:18:43.920 --> 00:18:59.759 So I'd frame it as a defensible option, supported by consistent exploratory data rather than something proven by a head-to-head test, and I'd weigh the tolerability cost against the expected gain for an individual patient kind of sitting in front of me at this point. 00:19:00.160 --> 00:19:00.640 Thank you, Dr. 00:19:00.720 --> 00:19:00.960 Moore. 00:19:01.119 --> 00:19:01.279 Dr. 00:19:01.359 --> 00:19:03.519 Askander, what is your approach to this? 00:19:03.680 --> 00:19:10.319 Have you started using Durvalimab elabrib combinatorial strategies in selected PMMR patients? 00:19:10.880 --> 00:19:16.000 Yeah, this is it's an important and nuanced uh question, both in decision. 00:19:16.400 --> 00:19:26.400 I have had a patient actually who had a known germline BRCA1 mutation and developed a metastatic serous endometrial cancer. 00:19:26.640 --> 00:19:32.240 And interestingly, her serous endometrial cancer was her 2-1 plus by IHC. 00:19:32.400 --> 00:19:41.200 And so when I was caring for this patient in my clinical practice, it was actually shortly after the time that the data was initially presented and published. 00:19:41.920 --> 00:19:43.359 We spoke about these results. 00:19:43.440 --> 00:19:50.720 There's, of course, strong biologic rationale for the use of PARP inhibitors in BRCA mutated ovarian cancer patients due to the amazing work that Dr. 00:19:50.799 --> 00:19:55.359 Moore, DiSilvestro, and colleagues did in the Solo 1 clinical trial. 00:19:55.519 --> 00:19:57.599

And so we spoke about the implications of that. 00:19:57.759 --> 00:20:08.240 We also spoke about the fact that they looked at BRCA and HR mutations as part of the clinical trial, but it was a small subset of patients and very difficult to understand the implications. 00:20:08.400 --> 00:20:14.319 But nonetheless, after our conversation in that situation, I offered it to this patient, and that's the decision we made. 00:20:14.400 --> 00:20:25.759 So she had chemotherapy and immune checkpoint inhibition, followed by immune checkpoint inhibition plus an oral PARP inhibitor as a maintenance strategy, and actually did very well for a period of time. 00:20:26.000 --> 00:20:34.559 Was on her maintenance PARP inhibitor for approximately, I want to say 13 months with her checkpoint until she developed disease progression. 00:20:34.799 --> 00:20:40.880 So clinically, I was hoping for a longer disease-free interval, of course, as we all do for our patients. 00:20:41.039 --> 00:20:46.400 But I did feel like in that scenario, there was clinical equipoids and having that conversation offering. 00:20:46.559 --> 00:21:00.960 Absent that population right now, I'm not routinely using a PARP inhibitor maintenance strategy in these patient cohorts or patients specifically who have, I should say, OLAPRA maintenance strategies in these endometrial cancer, PMMR endometrial cancer patients. 00:21:04.960 --> 00:21:06.480 And this is compelling data. 00:21:06.559 --> 00:21:13.119 There's no doubt about it, even if it is not powered and wasn't the primary question that the trial sought to answer. 00:21:13.200 --> 00:21:14.319 It's it's compelling. 00:21:14.480 --> 00:21:22.559 And we have patient populations we treat that with very high-risk histologies that are difficult to salvage. 00:21:22.880 --> 00:21:32.240 And I think when something's biologically plausible and there is evidence through shared decision making, certainly there are opportunities to offer more personalized approaches to their care. 00:21:32.400 --> 00:21:37.039 And hopefully we'll gain you know further insight into this with future trials. 00:21:37.599 --> 00:21:37.839 Dr. 00:21:38.000 --> 00:21:43.680 Eskander, are there any additional study considerations that our podcast listeners should keep in mind? 00:21:43.759 --> 00:21:44.160 And then Dr. 00:21:44.240 --> 00:21:49.519 Moore, I'll let you have the final word on any additional key takeaways you'd like our listeners to know. 00:21:50.319 --> 00:21:54.400 Yeah, I think I kind of began the conversation by mentioning this. 00:21:54.559 --> 00:21:59.759 I think the most meaningful contribution in this work is that. 00:22:00.079 --> 00:22:06.000 Biomarker analysis, the value of it is not necessarily only in finding a single predictive biomarker. 00:22:06.240 --> 00:22:11.200 It's really illustrating that this PMMR patient population is heterogeneous. 00:22:11.440 --> 00:22:17.440 And that most patients were identified to have a biomarker, as already alluded to, over half had more than one. 00:22:17.680 --> 00:22:37.359 And what I love, and I and I talk to this about the fellows of my training program and colleagues, as I say, we're living in a time where drug development is happening at a pace that is incredibly rapid, and the technology is available, and being able to design therapeutic drugs that are able to target a particular biomarker is continuing to emerge. 00:22:37.519 --> 00:22:38.079 You know, Dr. 00:22:38.240 --> 00:22:57.920 Vader, you yourself are now leading an NRG cooperative group trial that's trying to take advantage of the P53 molecular aberration to prospectively validate that that is going to be a biomarker that's going to help us in inform treatment strategies with the addition of anti-angiogenic therapy, potentially with immune checkpoint or bibicism ab on its own. 00:22:58.000 --> 00:23:00.160 So I am super excited about this. 00:23:00.319 --> 00:23:12.720 Again, I think this is critical for us to do for more of our studies because really this is going to be the foundation for which we move the science forward and begin to continue to identify therapeutic options that most likely will benefit our patients. 00:23:13.039 --> 00:23:13.599 Thank you, Dr. 00:23:13.759 --> 00:23:14.319 Eskander. 00:23:14.559 --> 00:23:14.720 Dr. 00:23:14.880 --> 00:23:15.359 Moore? 00:23:15.920 --> 00:23:18.480 Yeah, I really agree with all of those comments. 00:23:18.640 --> 00:23:25.680 I would maybe just come back to closing on DuoE and what we've learned from this particular analysis. 00:23:25.920 --> 00:23:37.599 It adds to the literature, you know, from the studies that Ramez and Matt Powell and others now you have led, that PMMR into mutual cancer is probably a term that will not need to go away. 00:23:37.839 --> 00:23:40.880 Probably is going to be not DMMR in the future. 00:23:41.039 --> 00:23:45.759 It's a very molecularly heterogeneous group with a heavy biomarker overlap. 00:23:45.920 --> 00:23:51.680 And I think that's kind of the headline for this podcast and will be the headline for how we design studies moving forward. 00:23:52.160 --> 00:24:06.480 Specific to DuoE, the benefit of adding Olaparib, we think was broad across the biomarkers that we evaluated in histologic subgroups rather than restricted to a single molecular niche, which was actually what I thought was going to be the finding when we did this. 00:24:06.559 --> 00:24:11.119 So that's why we do studies, because often you're wrong when you have a bias. 00:24:11.440 --> 00:24:20.079 So I think it does present some support for a combination strategy over a narrow biomarker-gated sort of strategy. 00:24:20.240 --> 00:24:30.640 And they do support use of this combination as a potential new option for PMMR disease, which would expand on the primary analysis. 00:24:30.799 --> 00:24:33.119 But what we are waiting for, and Dr. 00:24:33.200 --> 00:24:37.119 Weston will present hopefully in the near future, is the mature overall survival. 00:24:37.279 --> 00:24:44.160 So I think that's really going to be quite important for putting any PFS advantages into context. 00:24:44.559 --> 00:24:52.240 More prospective validation of these exploratory biomarkers, as uh Rama's already alluded to, many of them are not validated in this setting yet. 00:24:52.480 --> 00:25:06.720 Building HER2 and other emerging targets into future correlative work, and then doing all this molecular classification up front is what's going to make any of this actionable at the bedside, you know, in real time for patients moving forward. 00:25:06.880 --> 00:25:12.000 So lots learned here, but still lots to do. 00:25:12.480 --> 00:25:24.319 Well, thank you both so much for being on our podcast today and for your remarkable research contributions to our field that are truly improving or saving many more lives. 00:25:24.480 --> 00:25:33.599 And this was a remarkable and provocative work that answered many important questions and leaves many questions unanswered, but it's exciting to look ahead at what's to come. 00:25:33.680 --> 00:25:37.200 And I suspect those questions will be answered in the coming years. 00:25:37.519 --> 00:25:47.839 This article and the editorial we have discussed are available at gynecologic oncology dash online.net, and you can access our podcast at the journal's homepage, Apple iTunes, Google Play, and Spotify. 00:25:48.079 --> 00:25:49.839 Thank you so much for listening.