Good morning, and welcome to the ImmunoGen SGO Investor event. Today's conference is being recorded. At this time, I'd like to turn the call over to the ImmunoGen team to get started. Good morning, and thank you for joining us today, particularly those of you here in the room with us in warm and sunny Phoenix. Today's discussion will include forward-looking statements. These statements involve certain risks and uncertainties that could cause our actual results to differ materially and speak only as of the date of this presentation. Refer to our forward-looking statement slide in the accompanying deck and the risk factors outlined in our SEC filings for additional detail. In just a moment, I'm going to hand the call over to Dr. Anna Berkenblit, our Chief Medical Officer, to host a panel discussion with our guest today, Dr. Ursula Matulonis, Chief of the Division of Gynecologic Oncology at the Dana-Farber Cancer Institute and Professor of Medicine at Harvard Medical School, and SORAYA Co-Principal Investigator. Dr. Robert Coleman, Chief Scientific Officer of US Oncology Research, also SORAYA and Co-Principal Investigator, and Dr. Gottfried Konecny, Professor and Lead Clinician for Gynecologic Oncology in the Department of Medicine at UCLA. Just a few comments to start. Bringing mirvetuximab to patients with platinum-resistant ovarian cancer is our highest priority. We were thrilled to have Dr. Matulonis present the full results from the pivotal SORAYA trial yesterday afternoon in a plenary session at the SGO annual meeting. Having engaged with cancer experts over the last few days, the enthusiasm for these data and the broader mirvetuximab program is high. We see this as an important moment both for the company and for patients. For ImmunoGen, the SORAYA results are the culmination of many years of work dedicated to developing the next generation of ADCs to improve outcomes for patients living with cancer. We're on track to submit the BLA for mirvetuximab before the end of this month. We look forward to bringing mirvetuximab to market upon approval, transforming ImmunoGen into a fully integrated oncology company. For patients, the SORAYA data represent a significant step forward in the biomarker-defined treatment of ovarian cancer. We are seeking approval for mirvetuximab for second through fourth line patients with folate receptor alpha high platinum-resistant ovarian cancer who've been previously treated with bevacizumab. This would be the first FDA approval in over seven years, specifically for patients with platinum-resistant ovarian cancer. We believe mirvetuximab has the potential to transform the treatment paradigm to become the new standard of care in this setting. Exciting times ahead, and today we're pleased to offer the opportunity to share insights from experts in the field. With that, I'll turn the call over to Anna, who will moderate today's discussion. Anna? Thanks, Mark. I'd like to begin by asking Dr. Matulonis to summarize the full SORAYA data that she presented yesterday beautifully. This is really to level set and set the stage for a lively conversation with all three of our panel experts today. Over to you, Ursula. Good morning, everybody. I'm just gonna start with the background slide. We know that treatment options for platinum-resistant ovarian cancer are very limited, and they consist of single agent chemotherapy, which have response rates in the single digits, typically. Many of our patients will already have received bevacizumab, either as part of their upfront regimen or in recurrent setting. There is no known biomarker-directed therapy, that's indicated for patients with platinum-resistant cancer. Ovarian cancers, specifically high-grade serous, overexpresses folate receptor alpha. Overexpression of FRα is associated with poor clinical outcomes. We know that 40% of high-grade serous cancers have high levels of folate receptor alpha expression. Mirvetuximab soravtansine is a first-in-class antibody conjugate comprised of an FRα binding antibody, a cleavable linker, and a maytansinoid DM4, which is a potent tubulin targeting agent. The pooled analysis from previous studies of mirvetuximab identified 70 patients with FRα high platinum-resistant ovarian cancer, 1 to 3 prior lines, all with prior bevacizumab. The overall response rates are listed on the slide, 31% response rate. That data really serves as the baseline in how this trial was developed. SORAYA is a global single-arm phase III study evaluating mirvetuximab in patients with FRα high platinum-resistant ovarian cancer, and that also includes primary peritoneal and fallopian tube cancers. 106 patients were enrolled in the study. 38% had stage four disease. A bit higher than we usually see, but I think that speaks to the aggressive nature of the cancers. There was also a heavily pretreated population, and 51% of the patients had three prior therapies. All patients received bev, and that was a requirement of the study, and 40% of patients had received a prior PARP inhibitor. Where are we? Yeah, next slide. The primary endpoint of the trial was confirmed overall response rate. Am I on the right slide? Mm-hmm. The primary endpoint of the trial was confirmed overall response rate as assessed by the investigator, and this was 32.4%, including five complete responses, which we typically do not see in this population. The response rate far exceeds those seen in prior studies of less heavily pretreated patients with platinum-resistant ovarian cancer. The overall response rate by the investigator and by blinded independent central review, or BICR, were comparable. You can see that on the slide. Importantly, the median duration of response was 6.9 months by the investigator in how we assessed the patient's response. Importantly, these data are consistent regardless of the number of prior lines of therapy or prior PARP inhibitor use. You can see that here. Mirvetuximab's safety profile is characterized by predominantly low-grade, all reversible ocular, and they had low-grade GI events as well. These toxicities are generally managed with supportive measures and dose modifications if needed. The most common adverse events were blurred vision, keratopathy, nausea, dry eyes, and fatigue. The rate of patients discontinuing for treatment-related adverse events was low, with only 7% of patients discontinuing treatment. Treatment-related adverse events led to dose delay in 32% of patients and dose reduction in 19% of patients. Ocular events are common with antibody conjugates. However, we all know that ADCs are not all the same. The ocular events observed with mirvetuximab primarily included low-grade blurred vision and keratopathy, which were managed with dose modifications and were removed. All patients used prophylactic lubricating and steroid eye drops, which were mandated by the protocol, and all patients had a baseline ocular exam. Those with ocular symptoms underwent sub exams in two cycles, and patients were followed for their eye toxicities. Events were generally predictable, with symptoms typically occurring around cycle number two. In most patients, about 60%, symptoms resolved prior to their next cycle of treatment, and the remainder were managed with dose modifications. Less than 1% of patients discontinued therapy because of an ocular event. The one patient who discontinued had resolution of her ocular event in 15 days. There were no corneal ulcerations, and there were no perforations in this study. In summary, we believe these data have the potential to be transformative for ovarian cancer patients and their physicians in the platinum-resistant setting, and particularly in later-line patients. Response rates available with available therapy are in the single digits, along with significant toxicities. We are highly encouraged by the SORAYA data and believe these results show the potential for mirvetuximab to become the standard of care for patients with FRα-positive platinum-resistant ovarian cancer. With that, I'll turn back over to Anna. Great. Thank you, Ursula. Welcome to the panel, Rob and Gottfried. To begin with, I just want to add a little bit of color around the SORAYA data before we jump into questions. You know, one question we've been getting is how to put the PFS data from the SORAYA trial into context. As you saw in one of Ursula's slides, the median progression-free survival in this heavily pretreated population is 4.3 months. The first point that I want to begin with is that the antitumor activity of mirvetuximab has been consistent. With SORAYA, we've essentially replicated the data from that initial pooled 70-patient cohort back a couple years ago in terms of overall response rate, duration of response, and also progression-free survival. Given this, we believe that the MIRASOL data similarly will be consistent with the prior results from FORWARD I. As you can see here in this slide, the median progression-free survival in FORWARD I for the FR alpha high patients via the PS2+ scoring, all platinum-resistant, one dose required, was 5.6 months, both by investigator and blinded independent review. The hazard ratio was 0.55-0.62 compared with chemotherapy, demonstrating a remarkable treatment effect for mirvetuximab over and above what you can expect with available therapies. Given the similarities of the population between FORWARD I and MIRASOL, we anticipate a similar treatment effect in MIRASOL, noting that we've also designed it more conservatively around the hazard ratio of 0.7. Thus, we believe that the MIRASOL data have the potential to show a median progression-free survival of around 5.6 months as well. Now, one question that people have raised regarding any differences in the population from FORWARD I and MIRASOL is, what about PARP use? When we enrolled the FORWARD I study, PARP had not yet been integrated into the treatment landscape, so only about 10% of patients had a prior PARP inhibitor. We now know more and more patients are appropriately getting a PARP inhibitor, and in the SORAYA study, almost half of the patients did. We have a nice population where we can say, how does prior PARP use affect mirvetuximab activity? The answer is it doesn't. We have very nice activity from mirve regardless of prior PARP use or not. This gives us even more confidence, if anything, in the probability of technical success with MIRASOL. With that, I'd like to bring us back to the SORAYA study. I'd love Ursula, Rob, and Gottfried, each of you to just reflect on what the SORAYA data mean for your patients. You know, as I mentioned before and mentioned in the slides, our patients with platinum-resistant ovarian cancer, once they've received bevacizumab, have very, very few agents available to them. They have single-agent chemotherapy, such as drugs like topotecan and irinotecan. These drugs have very low single-agent response rates, you know, typically less than 10%. Patients have pretty quick progressions. Unfortunately, they pass away. There hasn't been any new treatments really in sort of just under a decade, and there's certainly no biomarker-directed therapy that's available for our patients. I have many patients who are waiting for this drug to be available, because of the activity where we can select patients whose cancers are FR alpha high, and give them the drug where there's responses. The toxicity profile is very well-tolerated. I've treated many patients with this agent. There's no alopecia. You don't have to fool with the Paxman cap. They can really not worry about losing their hair. The eye's toxicities are reversible. That duration of response is close to seven months. It's an impressive drug, and again, as I said, patients are waiting for it. The only thing I'd add to that is that if you haven't been in a clinic with platinum-resistant ovarian cancer patients, you don't really understand the, you know, what the natural history of this is. It's almost. It is very much the case that we're going from drug to drug to drug to drug until it exhausts all the options. Once a patient becomes platinum-resistant, there just aren't that many options. If you look at the compendia listing, we have a lot of drugs that have been studied, but they don't work. Most of those have not been superior. All of our head-to-head trials have been negative. It's an enormous benefit to marry the biology that we have come to understand in recent years with an agent that is targeted towards it and works. Patients will hunt down the access to this drug because they know that the available options are so limited. Right. But it's the disappointment on bringing a CAT scan back to a patient after two cycles of therapy and knowing you have to change again is just devastating. It is so important for us to continue to push the needle, and that's why we're so excited about this. I mean, it's just. There's just no other way to say it, but you've got an opportunity. Thanks for highlighting the unmet need, Rob. Mm-hmm. Gottfried. Yeah. You've had a lot of experience now with mirvetuximab. Yeah. I've been taking care of women with ovarian cancer for about 25-30 years, and every couple years, there's an exciting drug that comes along. Mirvetuximab soravtansine is definitely one of those for me. Rob also described the role of crossover patients with ovarian cancer. You have, you know, short responses, brief successes, and then they tumble from one treatment into the other. At this point, with platinum-resistant disease, we have a paucity of options to offer. They're very toxic. They're only a short duration, and there's a real unmet need for that. My experience has been having a drug that has such a good toxicity profile, no alopecia, incredibly important for patients that have lost their hair already twice in their lifetime. Manageable eye toxicity, very important. We have to differentiate mirvetuximab from other ADCs, which we've been exploring in many clinical trials. This is manageable. Also the durability of its efficacy, it's for me, astounding. I have patients that have been on the drug for two and a half years, that's unthinkable with another drug. I've been very excited about this early on. SORAYA confirms this experience. I think of many patients that sense this disappointment when they're in a randomized trial and they don't get the actual treatment arm. That's not because we bias them towards the active drug. It's because they come with an expectation to have a targeted therapy. You select them based on a biomarker. We're living in 2020. Patients come to you for rationally designed treatments. If you provide an all-comers typical chemotherapy drug, there's disappointment. They've been through that two or three times. There's an incredible selling point in providing an ADC that's effective, that has a rationale behind it. For me, this is a no-brainer. Mm-hmm. Honestly. Sorry. Thank you, Gottfried. That's a nice reminder that we do have a randomized trial ongoing with MIRASOL. For patients now who are aware of the SORAYA data, they would also be eligible for the MIRASOL study. That MIRASOL study is designed and intended to be the confirmatory study. Rob, given your. Mm-hmm. Experience with FDA, you know, there's been a lot in the news lately focused on accelerated approvals. Would it be helpful if you could provide a little color from your perspective? Yeah, there has been a lot of attention, you know, and even recent congressional activity with respect to the Accelerated Approval Program, which is set to expire in October, or to be reviewed for renewal in October. The one thing, though, I will say is that the current team at the agency is a very decently experienced group with ovarian cancer. That had been our struggle before, was getting them to understand, you know, what was clinically relevant. They have provided us with consistent guidance. I'm comforted this year that they have opined in similar way, and across our experience with interactions with them. We have a paper coming out very soon to talk about endpoints that we wrote with them. We'll provide some more guidance with respect to our recent interactions. It's been good. Great. Thanks, Rob. Now I'd like to turn to this being a biomarker-selected population. You know, 10 years ago in ovarian cancer, we were just trying to get people tested for BRCA mutations from a germline perspective. Now we have somatic testing, we have HRD testing. Can you talk briefly about how you incorporate the available testing now and how and when you might test for folate receptor alpha in the future? If we could go left to right. Oh, sure. Gotcha. Oh, sure. Gotcha. You know, we know that the levels of FR alpha expression don't significantly change between newly diagnosed state and then recurrence. That the archival tissue from someone's initial diagnosis when they go to surgery, then we have pathology stored tissue available that can be retrieved and then stained. You know, I think as oncologists taking care of patients with ovarian cancer, we're very used to now ordering tests. We'll order certainly Next Generation Sequencing. We'll order obviously BRCA status, both somatic as well as germline testing. I think the bottom line is people are used to ordering tests, and we will be able to order FR alpha and see what the level of expression is to make a determination if the patient's appropriate for mirvetuximab soravtansine. I think that's now, you know, ingrained in us to do that. That's not gonna be, now we're ordering for two uncertain patients, obviously not ovarian cancer patients, but certainly other patients. I think that's gonna be easy to do. I think the availability and cost considerations for comprehensive testing is becoming more available and affordable to patients. I think and what's happening is that we're trying to stuff in as much as we can so that the one-time evaluation will provide us with a lot of information. We've already been meeting with companies to figure out ways to get this into the panel. I think tailoring treatments and precision medicine is the future. If you want to invest in future this is the way to go because, cost containment, we can't provide the same treatment to all patients without pre-selecting them, those that benefit the most from this treatment. Having someone who left Europe 20 years ago to develop precision medicine in breast cancer with UCLA's group, such as Herceptin, we've lived through these examples in breast cancer, selecting a subset of patients at breast, only 20, and everybody knows the success of this drug. It's taken too long to move this into GYN oncology. I see, these are really the initial steps, moving towards personalizing treatment. I've asked our pathologist long time ago, "Can you add folate receptor testing to the panel?" It's not a problem because we already do 10 IHCs using PAX8 and et cetera, other markers. This is easily implementable as an early test and should become standard like BRCA testing, in my opinion. Thank you. Turning toward the future, thinking about the MIRASOL study. You know, I don't know if it's folklore or a piece taken out of a guidance somewhere saying that, you know, you need to have a three-month improvement in progression-free survival for it to be clinically meaningful. I think it might be helpful for us to think about that in the context of the population, the patient population. Certainly, someone's progression-free survival with earlier stage or earlier breast cancer, you know, 12 months PFS, you would want a three-month improvement in progression-free survival to be clinically meaningful. Can you all, from your experience taking care of these platinum-resistant ovarian cancer patients, educate us a little bit about what would be a meaningful improvement in progression-free survival and how you put that into the context of the totality of efficacy data, rapidity of response, duration of response, duration of stable disease, all of that. How does that inform your treatment decision-making? Maybe we can just, yeah, whoever wants to go. I mean, everybody on this panel, also Rob. Yeah. We know the difference between statistically significant and clinically different or clinically meaningful. Rather than following in on an is it three months, is it four months, or is it six months discussion, yes, clearly, there is a cutoff where you could say there is a clinical benefit. Beyond that, it's also the incredible lack or absence of toxicity, which we lose out of sight when we just look at response rates and progression-free survival. Even if this were a drug to be equivalent with a two to three-month improvement in progression-free survival, patients will be asking for it because the options are liposomal doxorubicin, topotecan, losing your hair, maybe a drug that's approved in lung cancer, and then you start scratching your head. It's an unmet need. We need more drugs for patients with platinum-resistant disease. Personally, I'm not looking for a benchmark. Clearly, you need statistics to show that you're not being misled, that there is superiority. I wouldn't evaluate a drug solely on, oh, it just made six months. I don't know, Rob, what you think. Oh, my gosh. Yeah. It makes me want to go off the deep end on this. Your question was really interesting because you brought in a population element and you brought in a patient element, and they're just different. If you think about it, your question about where it came from. The reason where it came from is that when we provided guidance to companies, when they are setting up their trial for statistical meaningfulness, we set up that three-month difference, which essentially is 2 assessment cycles different. Whatever, if it's 6 weeks assessment or 8 weeks assessment, we basically try to build in a 2 assessment difference at the median. That's how we do it. We back math it and come up with what the hazard ratio should be. It's. We want it to target somewhere in that 0.6 range. That's where that all makes sense. That's where we start. That's population-based. You asked the question about the patient. You mentioned about the patient. Mm-hmm. The patient comes to the office, and they say, "What should I get?" I say, "Well, I can get you this treatment, and this is what I think the efficacy is gonna be." What's the efficacy? Well, the efficacy is response and delaying progression, okay? 'Cause that's what we're gonna delay. Then she says, "Well, how well does this work?" Well, I can quote you what the statistics is for the statistic, but what's the probability that they're gonna progress? Okay, that's not the median. Mm-hmm. It's the hazard. I can't drill this into you guys enough. It's the hazard for the event. It's what is the risk of a progressive event tomorrow? What is it gonna be next week? What is it gonna be in three months? Okay, that is an average risk that is measured by the hazard function. If there's a hazard of 0.6, what does it tell you? It means that every point of exposure, there's a 40% reduction in the risk that that patient's gonna suffer a relapse. It drives me crazy to talk about medians. It's a midpoint on a morphologically dependent curve. If the curves look like bananas, it could be a big spread. If they look like Ls, they could be exactly the same. What does it matter? It's the risk reduction. When a patient comes in for cycle number two, there's a probability that that event didn't happen, and it's been benefited by the effect of the treatment. That's what you should be focusing on. Sorry, I gotta move on. No. Thank you. Anything to add, Ursula? Well, no. I was just. I, you know, I concur with both Gottfried and Rob, and I think that certainly the PFS of 4.3 and then obviously the bigger of 5.5 months, I think is clinically significant. I mean, it certainly is statistically significant, but I think from a patient standpoint, to put her in remission for that much longer, using a drug that has, you know, totally acceptable side effects, you don't get the rashes with, like you see with pegylated liposomal doxorubicin with weekly paclitaxel. The level of bone marrow suppression is extremely well-tolerated. I can't emphasize enough the tolerable side effect profile of this agent. Of course, there's great response rate and, you know, 6.9 months of duration of response. When we came here, we talked about, you know, exchanging some experiences on patients, and we both thought that the durability is something. When you treat a patient with any of those drugs that you mentioned, for example, retreating them with a similar drug to the payload of this, like Taxol. Mm-hmm. Patients develop nail changes. Mm-hmm. neuropathy, not to mention the alopecia. When they're on it, at least I mean, hardly can treat someone longer than six months. Right. Yeah. Right? Very poor quality of life. Having absence of that, delivering the chemotherapy, in a more precise way reduces this toxicity, and that's already for me a convincing point. Yeah. The other thing that I just to highlight is that this is something that where the responses are noted actually pretty quickly. That's what's really encouraging. When a patient comes back for their first assessment, and they get a CA 125 or they get a CAT scan or whatever, a high proportion of them have already had the response or showing a demonstration of response or lack of progression. Yeah. That's so encouraging for our patients. I was gonna add to that. Yeah. We've had a few patients who've been just, you know, super sick. Yeah. -with, uh, with- Right. the burden of cancer, and they've really sort of been resurrected Yeah. Yeah. by this drug Absolutely. because of those impressive response rates and deep responses. Thank you. You know, Rob, you mentioned CA 125. Mm-hmm. CA 125 is a pretty good marker to follow the activity of a patient's disease if it's elevated. Mm-hmm. When we know they have disease, can you, all three of you, talk a little bit because, you know, those of us doing clinical trials, we need to have confirmed responses per RECIST, which is this very clear and arbitrary line. Mm-hmm. That's not what happens in the clinic when you see a patient. Can you talk about CA 125 and, like, how you think about the benefit for patients aside from confirmed response? Well, I can start on this one. The CA 125 is just, as I tell patients, one of the kind of components of how I assess how that patient is doing. Not infrequently, the CA 125 will go in one direction, yet a scan might go in a different direction. I think what ends up happening in patients with platinum-resistant, more heavily pre-treated cancer, there are differential sites of disease. There can be nodal disease, there can be peritoneal carcinomatosis, there can be ascites. There may be, for example, she gets better because her ascites dries up, her peritoneal carcinomatosis gets better, but that may be a lymph node gets- Yeah. a little bit bigger. On study, that might indicate progression or a new lymph node comes up- Mm-hmm. that's tiny but new. I think in the real world practice- Yeah. You know, we use clinical exam, how the patient's feeling, CA 125, and then scans to kind of pull together. Mm-hmm. that assessment of the patient. Yeah. I think, Anna, you've mentioned a good point. It does for people that have high dynamic ranges in CA 125, that can be very predictive of what's going on. But on my personal opinion, if I get something that's keeping a patient from progressing, I'm gonna keep them on no matter what 125, to be honest with you. But it does in some it's nice for the patients, when, you know, when they think about it, they're so addicted to the CA 125. We started it from day one. They love to see the fluctuations going down. When it goes up, you know, I still will evaluate before I start approaching. You're right, we have a lot more flexibility in the real world than we are on a clinical trial. You know, we may even see better results. Yeah. Definitely. Mm-hmm. I agree. CA 125 loses its accuracy the further along you get in treatment and because of the mixed response that Ursula Matulonis just mentioned. Thank you. We've submitted an abstract to ASCO with additional efficacy data from the SORAYA study. We look forward to sharing that in a few months, hopefully in person. With that, I think maybe we can turn it over to questions from the room. Sure. Hi, good morning. Thanks for holding the event. I'm John Newman from Canaccord Genuity. Just had a couple of questions here. I guess maybe a question for Anna and maybe also for the panelists. This has actually just been discussed. Investors have been very focused on median progression-free survival from SORAYA to inform potential results from MIRASOL. We actually just heard Dr. Coleman talk about hazard ratios and what that means. Anna, I'm just wondering if you could talk to us about how the shape of the Kaplan-Meier curves is important in terms of showing a benefit? The reason I'm asking, I'm thinking about the difference between the PS2 FR alpha high analysis from FORWARD I versus the original analysis for FORWARD I, which was 10X. I'm just curious if you could talk about that, 'cause I know we're all very focused on sort of this one time point. Just curious if you could talk about the curves. Thanks. Thanks, John. For those of you who attended the Neil Love ISS yesterday, Rob and Ursula did a grand job there. Rob, I believe you presented one slide with the PFS data from the ITT population and the protocol-specified 10X FRα high. There was definitely a treatment effect in that 10X FRα high. The PFS curve though, it looked a little bit like PFS. Yeah, step. Yeah. Yeah. There were a couple points where the curves touched. Mm-hmm. They separated, and there certainly was a hazard ratio improvement, but wasn't a blazing signal. Mm-hmm. When we went back and we rescored by the PS2+, this is post-hoc, this is exploratory, but this is more of a blazing signal and exactly what you were describing, Rob, when you were describing the shape of a PFS curve. Mm-hmm. On this slide here, you can see this was a post-hoc exploratory analysis from FORWARD I, FRα high patients. You can see that the curves separate early to the points that you were making, that two-thirds of the responses occur within the first two cycles. You start to see benefit from mirvetuximab over and above the available therapy early on, and then the curves stay separated. Then there's a tail on the curve. Now, admittedly small numbers, but tail on the curve. Each of you with experience with mirvetuximab has had those patients who've been out there for one years, two years, three years. I think we have one patient on for four years. Mm-hmm. That really tells you the treatment effect of mirvetuximab above investigator choice chemotherapy. That median is just one point on the curve. Mm-hmm. At every point on the curve here, you can see benefit from mirvetuximab reducing the risk of progression exactly as Rob was describing. Actually, when I saw the response rate and the duration of response, I actually predicted what the PFS would be because of that curve. If you notice there, the first step down in the control arm is at 40%. That's very typical for recurrent, platinum-resistant ovarian cancer. 40% of patients progress at the first CAT scan. That's it. The rest of what's out in that tail now is gonna be made up of the responders, and we know what that duration is. Then it's made up of stable disease, which on average is usually for recurrent, two assessments. Do the math. You come out exactly to what we found out. I think that what you're gonna see here is the blunting of the stairsteps, and that's what you know, we hope to reproduce this. Absolutely. Great. Thank you. Just maybe one additional question for the panelists and also for Anna. One of the things that I noticed. Just a little louder, John. One of the things that I noticed during the presentation yesterday was there's a difference between the investigator-assessed response and the BICR response, which by the way, has been seen with PARP inhibitors and other drugs in ovarian cancer. Wondering if you could talk a little bit about that, because in your PS2+ analysis for FORWARD I, they're quite similar, and that's obviously a randomized study. I know investors have been asking me, and they're just curious, you know, why is there this difference there? Thanks. Yeah. At a high level. Mm-hmm. Across a large randomized study, there is an expectation that if the study is well conducted, there will be concordance between the blinded independent central review and the investigator assessment on a population basis. Mm-hmm. On an individual patient basis, there will be some differences. You just heard from all of our panelists that it's not just about the CAT scan, it's about the CA 125, it's about how the patient is feeling, the symptoms, what's on their exam. Investigators have a lot more information available to them when they are assessing whether or not a patient is responding or progressing. Mm-hmm. Now, when we're. You know, we have very nice concordance in our. Mm-hmm. Response. That's because response is really defined by this 30% decrease in the sum of longest diameters. Patients will be feeling better if their tumors are shrinking. That concordance makes a whole lot of sense. It's a little more challenging when you think about declaring when a patient progresses. Duration of response is a function of they're responding until they've progressed. Progression is defined as at least a 20% increase in the sum of longest diameters from the nadir, from the smallest amount. Mm-hmm. That's a little harder to say when. Mm-hmm. Dr. Matulonis. Yeah. Dr. Coleman, Dr. Konecny, when you're seeing a patient, if you know they're feeling well, you know, you're gonna look at the CAT scan with a certain lens. Mm-hmm. That the blinded independent radiologist won't have. That's one reason that there may be a difference in terms of when progression is declared. The other is that if you have patients with bulky disease, RECIST requires you to identify a certain number of target lesions at baseline to follow. Mm-hmm. You can't follow all of them. The investigator's radiologist might follow these, and the blinded independent central radiologist might follow slightly other ones. As you heard from Ursula, there may be a little difference among the target and the non-target lesions that again could affect exactly when the blinded independent radiologist versus the institutional radiologist declare progression. On an individual patient basis, there will be differences. Mm-hmm. When you're looking at a randomized study, the totality of the data with the treatment effect, all of those little nuances will be washed out. Yeah. In a smaller study like SORAYA that's single arm, you expect to have some differences. I'm not at all surprised that our blinded independent central review, progression-free survival and duration of response are longer. I would also point out that for this presentation, we wanted hot-off-the-press, mature investigator DOR data, so the data cut for that was March third. There was not enough time for us to get all the scans from all the sites around the globe into the blinded independent radiology facility for the radiologist to read, to send the results back to us to generate the data for this presentation. That 11-month. Mm-hmm. Duration of response is from the November data cut. It is highly influenced by a few patients. Mm-hmm. Far out on the curve. Yeah. Again, we will have updated data. It will come down, but it will be longer than the investigator. Okay. DOR. Okay. Does that answer your question, John? Mm-hmm. I mean, I'm not surprised Robert published that investigator-assessed PFS versus independent radiology review should be closer together if it's a frontline study. Mm-hmm. You don't need independent radiology review if you do a frontline study. In the recurrent setting, there can be differences. Actually, it's reassuring that the independent review actually showed longer. Mm-hmm. Which means that's actually the more objective assessment than the investigator who's biased using multiple variables, mixed response, non-target lesions. You know, you can be influenced and less objective than an independent radiology review. Actually, this data would be the independent radiology review. Yeah. I think, I mean, the other point I would just add is that CT scans sometimes underestimate that level of peritoneal disease. Yeah. That thickening that we can see. Mm-hmm. Around the intestines. Yeah. Where the patient's starting to have some symptoms is not gonna be called on RECIST. Yeah. We can get a sense clinically and then looking at our CT scan images, that something is starting to happen. Yeah. Yeah, exactly. To Anna's point, you know, the definition by RECIST, you can only have two targets in an organ. You know, when you have a liver metastasis, you could have three or four or five, and you can pick different ones to monitor. It can lead to some variance there. Mm-hmm. Andy? Sure. Thanks for taking my question. Andy Hsieh at William Blair. Thanks for all the perspectives, and thanks for holding the event. I think the first question I have is could you take us to the office where you're communicating with a patient deciding between approved single-agent chemotherapy and potentially go on kind of a risky clinical trial for mirve? Could you walk us through the process of kind of communicating. Sure. You know. I mean, I think, you know, all three of us have worked with this drug. I've worked with this drug since the initial phase I trials. I understand. I mean, it's the risk. I would take the risk out of that or remove that word. Yeah. Exactly. Yeah. Exactly. I would not tell a patient that this is a risky clinical trial. Again, I think I really need to, you know, impart that these patients know they're gonna die. They're gonna die. They've got platinum-resistant ovarian cancer. What they have on the table for options are pretty lousy single-agent chemotherapy that we know have response rates that are less than 10%, and pretty significant toxicities. Then they get the drug, and then they progress, and they feel poorly. If the question is around getting patients onto the SORAYA study, you know, the most important thing was making sure that they fit eligibility criteria. Mm-hmm. The platinum-resistant ovarian cancer is high-grade serous, and the test came back as FR alpha high. Mm-hmm. I had no problems in referring patients into the trial at all. Again, there's been just troves of studies published on this drug that de-risk it and make it a very well-tolerated agent. Yeah. I think. You know, at the patient level, when you're having the discussion, they're well aware of what their options are. I mean, they're not. They're not coming in with some unrealistic expectations. These types of trials are very easy to gain patient acceptance. That's just the bottom line. You gotta remember that when a patient comes in, if they're bothered by something, you know, there's a symptom that they're carrying, you know, half those patients or more will progress at the next CT scan. All of what they had before is gonna be carried to that point in time. Yeah, I think when you advise patients with recurrent platinum-resistant ovarian cancer, you offer them a number of options. You have to be honest to say that there is probably no major difference between drug A or drug B, as the available standard of care is not very effective. It really wouldn't matter if you pick drug A over B, because of the paucity of data. However, you then start to focus on quality of life. You leave the efficacy out of the picture. As long as you have an effective drug, you're not looking for a superiorly active drug. We don't have randomized data in the third-line platinum-resistant disease where we could choose one and advise a right or a wrong drug. It's really an individual decision to say which drug matches what patients expect, what side effects are they willing to endure, what have they suffered in the past. That's when the toxicity discussion comes in, lack of alopecia, lack of peripheral neuropathy. Being able to be on a drug for a long time, even if you don't obtain a CR or a PR. Those are all things that sway you and that make a patient interested. You know, I hope that it's very different in an early disease setting. In platinum-resistant recurrent ovarian cancer, you know, having a less toxic, equally effective, of course, if it's more effective even better, drug available, will be used. Agreed. One more, Andy? I have a question about so Dr. Coleman, I think your relationship with the FDA. I'm curious about their perspective on biomarker. I think biomarker is kind of a theme during the conference. Mm-hmm. Just curious about the agency's view, whether they share that excitement with investigators. Also, I guess during the last discussion, kind of an investigator assess versus BICR. Mm-hmm. I think there could be, maybe, hopefully I'm putting words in your mouth. Mm-hmm. There's an emphasis on BICR. Is that also the agency's view that they put more emphasis on BICR than investigators? Yeah. Good questions. The issue about biomarkers is mostly focused around predictive biomarkers. We have biomarkers that have prognostic value, which has provided some stratification for what a population will do in the presence or absence of a biomarker. When you talk about biomarker in that context like we are with this, we're talking about predictive biomarkers. The emphasis and the focus is there because our expectations are that if a biomarker is relevant, it's predicting its benefit, that the magnitude of effect should be different. It should be better. They're very interested in identifying predictive biomarkers to help essentially even reduce the sample size that's necessary to demonstrate a benefit. Those are very key pieces of it. Yeah, there's a lot of emphasis on identifying biomarkers and using drugs that are aligned with that. It's mostly because it allows us to differentiate from the noise. I was gonna mention the last discussion that, you know, we have done probably 24 randomized prospective phase III trials comparing agent A to B. There have been only two that were negative. I mean, two that were positive, but they were positive in the wrong direction. When you're at the floor of efficacy, the idea of another ineffective drug distinguishing itself is really hard to do. Using your biomarker to identify an active compound will, you know, obviously, you know, provide a better benefit for that going forward. With respect to BICR, in open label trials, BICR has been pretty much a requirement. In placebo-controlled trials, it's not a requirement, but they do like it for sensitivity analysis. They've been very consistent about that. One more question if you don't mind, Anna. There's a difference in number between investigator assessment and BICR by about 10. Is that what you're referring to? That numbers will come in after the data. No, no. What you've noted, Andy, is that of the 106 patients enrolled, 105 were efficacy evaluable by investigator. 95 were efficacy evaluable by Blinded Independent Central Review. How did we define efficacy evaluable? We defined it as all patients who have measurable disease at baseline. That's it. In the investigator's assessment, there was one patient who came in who, after the fact, we realized, they realized that the patient had non-measurable disease. Because overall response rate is the primary endpoint and you need to have measurable disease, they couldn't be evaluable by the primary endpoint, so they were excluded from the efficacy evaluable population. When the scans for all 106 patients were sent into the Blinded Independent Central Review, there were 11. 106 minus 95. There were 11 where the blinded independent radiologist just did not feel comfortable putting a caliper there and saying this is measurable disease. It could have been necrotic lesions or they could have focused on the ascites and missed the liver met, whatever. Those patients were excluded from the efficacy-evaluable population from BICR. This is really the most conservative way to define an efficacy-evaluable population. It is completely the right way to do it for a pivotal registrational trial. You will see that other studies are designed with different definitions of efficacy-evaluable populations. In some studies, you have to get to that first scan, and you must have at least one post-baseline scan to be considered efficacy-evaluable. Then you need to understand, well, if a bunch of patients are not efficacy-evaluable, why aren't they? Yeah, we have limited time remaining, so in our time lapse, we'll just take two quick questions from our covering analysts online. The first is from Michael Schmidt at Guggenheim. On the ocular AEs, how do these AEs compare to some of the ocular AEs that have been seen with other ADCs? Great. I'll ask Ursula to first comment, given her, your experience with mirvetuximab. Then Rob, given your recent publication with Tivdak, maybe you could add a little color. Gottfried, your experience. Yeah. Certainly, you know, as I presented yesterday, you know, the vast majority of the ocular toxicities are low -grade, and that's certainly the case clinically. Patients will present with a little bit of blurred vision. They may need some stronger reading glasses, and the majority of those will resolve by the next cycle. And if for some reason they don't, then we dose modify or dose delay. Sometimes we'll go from, you know, the initial, 6 milligrams per kilo down to 5. But again, the important point is they're all able to be, you know, sort of treated and mitigated. And I think a really important point is they're all reversible. You want to. Yeah. I mean, ADCs are different, completely different. They have different payloads. They have cleavable and uncleavable linkers. They are targeting different proteins. The toxicities that some of these agents have target different proteins that are expressed on the conjunctiva. You can't compare eye toxicity data from one agent to the other. It's important to point out that only 20%-25% of patients with mirvetuximab really require some action, meaning either holding the drug or delaying it for one- or two-week period. You do not need eye assessments every cycle. You do not need eye cooling caps. The list and the discussion would be going on long. I do think a eye toxicity discussion would be blown out of proportion with mirvetuximab. They do exist, but they exist with other drugs is my short answer. Yeah. I couldn't agree more. They're just completely different, and they're very drug specific. You know, the manuscript you mentioned just got published in Gynecologic Oncology, and it was there to provide, you know, it was focused on the Tivdak, and in the discussion you'll see how basically what Gottfried just mentioned is all these different effects can be differentially classified by the different ADCs. In that situation, the extremes of the drugs are very, very different in terms of the ocular event that you see. All right. Quickly, one last question from Kelly Shi at Jefferies. Is mPFS evaluation included for Accelerated Approval decision by FDA? FDA cannot interpret progression-free survival from a single-arm study because progression-free survival is a measure not just of drug activity, but the tempo of the underlying disease in that population. Progression-free survival can only be interpreted in the context of a randomized study, which is why it's the primary endpoint in MIRASOL. Certainly, we are sharing the PFS data and the overall survival data from the SORAYA study to FDA as part of the BLA, but that's not the data that will be considered from an efficacy perspective. Thanks, Anna. That's all we have time for. With that, I'll turn it over to Mark for some closing comments. Great. I'd like to extend my thanks to the panelists, particularly for your insights and passion with respect to the program and also our deep appreciation for the patients, the families, the caregivers, and the other investigators that have supported SORAYA and the broader mirvetuximab program. We look forward to keeping you updated on our progress. As I said, the BLA to the FDA is due to go in before the end of the month, and we'll keep you updated as we move through the Congress schedule through the back half of the year. Thanks everybody for your time today.