Mersana Therapeutics, Inc. (0M40.F) Earnings Call Transcript & Summary

Good afternoon, everyone, and thanks for joining us at the Morgan Stanley Global Healthcare Conference. I'm Michael, one of the biotech analysts here, and it's my pleasure to introduce the team from Mersana Therapeutics, starting with Anna Protopapas, CEO; Tim Lowinger, Chief Science and Technology Officer; Arvin Yang, CMO; and Brian DeSchuytner, SVP of Finance and product strategy. So just a quick reminder to everyone, this is a fireside chat today. [Operator Instructions] But before we get started, I just need to read a quick disclaimer. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. And if you have any questions, please reach out to your Morgan Stanley sales representative. So with that out of the way, Anna and I'll turn it over to you to make some introductory comments, and then we can jump right into the Q&A.

Great. Thank you, Mike, and thank you for hosting us today. On Friday, we made a very substantial disclosure on our lead asset UpRi. And I thought I would start by just summarizing the key observations around the data we disclosed. The first part of our disclosure involved an additional update on our ovarian expansion cohort. This was the first update -- this was the fourth update in the expansion cohort and we presented 75 evaluable patients and showed a very consistent and clinically meaningful profile. We saw a response rate of approximately 34% in the NaPi2b high patients, and those represent 2 third of the ovarian cancer population. Remember, this is a very heavily pretreated patient group where the standard of care is no better than 12%. Also, our thesis for a differentiated tolerability profile without the severe neutropenia, neuropathy or ocular toxicity remains intact in this updated data disclosure. That was the first part of our presentation. The second part involves the data that supported a modest dose change of 15% or less, that we believe, based on a substantial body of data that will reduce the very rare grade 3 AEs, particularly the rare pneumonitis while maintaining efficacy. And let me double-click on this and explain the data set that supports this decision. Between the ovarian cancer expansion called the dose escalation data as well as the smaller lung cohort, we have almost 200 patients treated on UpRi today. That data set allowed us to build a very robust population PK model, where we were able to correlate exposure with efficacy and exposure with safety. In the exposure -- the exposure range that we are looking at between 36% and 43%, we see that the relationship between exposure and efficacy flattens out, while the relationship between exposure and safety remains steep. This indicates that a modest dose reduction can sustain efficacy while improving tolerability. Once we had the population PK model, we looked at the patients we had dosed in the expansion cohort and really bifurcated those patients into those that received around 36 milligrams per meter square, the lower dose and those that received 43 milligrams per meter square at the higher dose. And we saw that the trends in those 2 different dose groups supported the observations from the PK modeling. We saw the grade 3 or above AEs really aggregate to the higher dose. We saw the higher dose have more dose reductions. We saw the higher dose have a shorter treatment duration and we saw the higher dose have more patients that discontinue early even before they are evaluable. We shared this data with investigators and it resonated with them because it was consistent with their patient experience. There are other ADCs that have shown that there is always the sweet spot between efficacy and safety. And I think with the substantial data we have, the power of 200 patients being treated on UpRi, we are able to really optimize the profile and find that sweet spot. The third part of our presentation involved UP-NEXT, a Phase III trial informed by FDA feedback that can service a confirmatory trial as well as the path for global registrations. UP-NEXT is a trial that would allow us to end to enter the platinum-sensitive space, a market opportunity many times larger than platinum resistant. We will be addressing unmet medical needs in the platinum recurrent maintenance setting, and the FDA has agreed that given the level of unmet need and the lack of standard of care for these patients, the trial can be done against placebo. So those were the 3 areas we discussed. Updated data, indicating a very consistent and robust profile and opportunity to optimize the profile with a modest dose reduction supported by very substantial data set and then a path not only took a confirmatory trial, but one that would expand substantially the market opportunity for UpRi. So maybe with that as an introduction, Mike, we would be happy to take any questions you have.

And maybe just to start, for people aren't that familiar with Mersana, if you can just talk about your ADC platform. You have a few different options there and maybe talk about some of the unique features of each or what's different between them?

Maybe I'll ask Tim to answer this question.

Sure. I'd be happy to. So thanks for the question, Mike. And you're right, Mersana is a platform-based company. And I think we have a set of innovative and highly differentiated technologies, which we've really developed to overcome the limitations of earlier platforms. For example, starting with our antitubulant payload, which we call DolaLock, it has been specifically designed to have what we call a control standard effect that we believe allows us to recognize efficacy, while avoiding the severe neutropenia, peripheral neuropathy and ocular toxicity that has limited many other anti-tubulin platforms. The DolaLock Payload is used in both of our clinical stage platforms in UpRi. It's in dolaflexin, which uses our biodegradable biocompatible polymer called Fleximer to create dolaflexin and allows us to achieve quite high drug to antibody ratios, such as an UpRi, [indiscernible]. With Dolasynthen, we've taken everything we've learned from dolaflexin and created a platform that is fully synthetic and gives us more precision because we don't believe it's one size fits all. And so with Dolasynthen, we can specifically create ADCs of various DARS, DAR2, 4, 6, 12, 18, make them homogeneous and identical, and we see some benefits of that. And that is the platform used in 1592 is also in the clinic. And then thirdly, there's a new platform that I'll be revealing more data on at the triple meeting later this year, actually next month. But we're very excited about it because it takes ADCs, we believe, firmly into immuno-oncology. It is built from the ground up, a STING agonist ADC platform, designed to increase the therapeutic index and activate the innate immune system. We specifically chose STING because STING is a fundamental immune activation pathway, and we've demonstrated that we can activate that pathway with this platform in both tumor cells and tumor resident immune cells, delivering what we call a one-2 punch. And we can do that with -- based on preclinical data with -- we're very pleased with the efficacy we see as well as the overall tolerability. And so those are the platforms that are either in clinic or close to the clinic that we think differentiates Mersana from other ADC companies.

Got you. That's helpful. And maybe just a quick follow-up for you, Tim. When you think about Dolasynthen versus dolaflexin and the ability to sort of fine-tune the DAR with Dolasynthen, how do you think about applying those 2 platforms? Are there particular situations where one platform might work better than others? Is it certain different indications? Or how you think about that?

That's a great question. I think it's one of the benefits of being a platform company. We've often said that it's not one size fits all. And so we're very data driven. And so for example, in XMT-1660, which is our B7-H4 ADC that is set to enter the clinic next year. There, we evaluated Dolasynthen DAR 2, Dolasynthen DAR 6, we also in head to head comparisons to dolaflexin of DAR 10 to 12 and let the data drive the decision. And that's what we do routinely. We think both of these platforms have differentiating factors. And we let -- rather than assuming one is most appropriate, we generate the data and based on that data, decide what -- has the most promise for patients.

Okay. Great. Thanks for that, Tim. So maybe we can just shift gears to UpRi. Anna, you did a nice job sort of giving us some updates there from last week, but maybe you could talk a little bit about the target NaPi2b and what makes it attractive, particularly for an ADC application?

So NaPi2b is a sodium dependent phosphate transport protein, and it's broadly expressed in ovarian, non-small cell lung cancer and some other tumors but has very limited healthy tissue expression. When we embarked on this program a few years ago, we had the benefit of some validation of the target that Genentech had done. They had taken an ADC using the vcMMA platform into the clinic and so activity but also saw a substantial amount of neutropenia and neuropathy, which prevented them for combining with other agents and moving into earlier lines of therapy. So as we embarked on the program, benchmarked our ADC against that earlier version, which has now terminated, as you know and we saw preclinically both improved efficacy and improved tolerability. So the Genentech validation, our head-to-head data gave us confidence that we had a good molecule against the validated target, and we -- and that's what brought 1536 now UpRi into the clinic.

Maybe we can talk a little bit about -- you mentioned this in your opening comments about lowering the dose for UpRi. And maybe you can talk about what the key trigger was in terms of that decision. I think you did a good job of mapping out the data you have that should support that, but maybe what was the trigger that sort of started that?

Yes. Why don't I let Arvin do that because this involved a very extensive amount of work from his group to analyze and really find the right path forward.

Thanks, Anna. So maybe just to start, the impetus in part to this was looking for an opportunity where we could address the rare grade 3 toxicities that were occurring. And so given that, as you can imagine, because they are so infrequent, we really needed to actually have a larger set of data in order to do a more robust analysis across the different dose ranges and so forth. And to that end, we had almost close to 200 patients in total that have been treated with UpRi. When you look at the expansion, the dose escalation, additional lung cancer patients and all of that, in aggregate, was utilized in order to power a robust analysis in relationship to our population pharmacokinetic modeling, which really looked at the data in relationship to on a continuous variable exposures and correlating that with probability of either higher grade toxicities of different manifestations or with efficacy. And the real take-home point here was that when you thought about that linear or continuous variable and the steepness by which the increasing exposures correlated with toxicity, that continued to increase between the 36% to 43% range, whereby when we looked at efficacy, as we've described previously, across the different doses, you're seeing a very flat or similar activity between the 36% and 43%. So that provided a significant opportunity by which we could identify and increase therapeutic index or an opportunity to optimize a reduction in -- or an improvement in safety, while still maintaining that consistent robust activity.

Got it. Makes sense. Maybe you can talk a little bit about just the UPLIFT study in terms of the design. I think there's some unique features there that you guys have included. Maybe you can just touch on some of those?

So I can continue on that from the standpoint of the design of UPLIFT, it really is an opportunity for us to have the potential for label differentiation and increase the probability of success because it's a 2 shot on goal strategy. But let me start with actually the patient population for a second. So we are allowing -- and this is all supported by the activity we're seeing in our expansion data up to 4 lines of prior therapies as well as actually, including both a pretreated as well as bev naive patients that are in a matter -- in a manner that's consistent with the bevacizumab label, meaning that only patients that have 1 to 2 prior therapies are obligated to have received bevacizumab and those patients that have more than that are not obligated. In addition to that, we're not excluding patients with underlying neuropathy, which can be a substantial patient when you think about where they are within their disease state and heavily pretreated. The other aspect to this is, as I alluded to, the 2 shots on goal. So we're allowing patients to come on unselected. However, once they're on study, we're identifying whether they are NaPi2b high or not. And our primary endpoint is in the higher NaPi2b patients and we'll be evaluating response rates. And that could support the potential that we have an indication that's within a higher NaPi2b population. But as I just stated, we're enrolling all patients. And so if the activity within the entire population is also supportive of a response rate that could support approval and I'll remind you that the standard of care within this population is 12% based upon single-agent chemotherapy data. And so if the second shot on goal is sufficient to support approval, then we would have an opportunity where we would have a broad indication, but still have our robust diagnostic assay available as a complementary diagnostic assay as appropriate.

Maybe just a quick follow-up there. You mentioned your biomarker in NaPi2b high patients. Can you just talk about how confident you are in that assay? And I know you guys have done extensive work in trying to define the cutoff, but maybe you can walk us through some of that.

Yes. We shared a webinar on this actually earlier this year in April. And just to give you the highlights of this. The development of the diagnostic assay was actually in parallel to the development of UpRi, even from the research phase. So it's been multiple years, and it's really resulted in what we believe to be a quite robust, predictive and reproducible assay that could result in a robust commercial potential diagnostic assay. So we're identifying patients that represent approximately 2-thirds of the ovarian cancer population built off of this assay. And we started with what was initially the H score, and we identified that at the cutoff of 110, it appeared to enrich for responses. But as we thought about evolving this to a diagnostic assay that could be easier and more readily adoptable, the H score, just to give you context, provide a wealth of information. It gives you not only the percentages of cells, but also the intensity of the cells in relationship with their positivity and it's well suited in relationship to the clinical research setting and for centralized meetings. But in contract as we thought about evolving this, we determined that actually the TPS, which is the tumor proportion of the score is a component of H core. And using that was actually adequate and sufficient and robust in relationship to and easier for pathologists to identify those same patients because TPS really was a strategy by end process by which pathologists have to essentially only distinguish between 2 colors, whether it be blue and brown, but not a gradation of colors in order to determine if a cell were positive or not. So built off of this, we believe that we've had this robust and reproducible asset that we'll be able to move forward. And further emphasizing where they actually complement where our technology lies is that we believe that based upon our ADC technology with the higher DAR or the drug to antibody ratio, each of these antibodies are delivering more payload and perhaps the intensity of expression may be less important. So the CPS, which is providing a proportion of positive cells regardless of their intensity, may be able to adequately and effectively predict for enriched responses.

Got it. That's helpful. Maybe just last question on UPLIFT. Can you remind us maybe when you started the study or how many patients have been enrolled so far? And how do you -- what impact could changing sort of the dose have in terms of the data analysis? And are there a couple of scenarios you can think of or share with us or if you only have to focus on low dose patients, does that mean you might potentially have to enroll sort of more patients to kind of maintain the sort of powering of the study?

Great question. So we have guided that we believe that during the summer of 2022, we should be substantially enrolled. And as for the patients that are already on trial, we have already sent the amendment out to the sites. It's a simple amendment, so it should be adopted reasonably quickly. The patients that are already enrolled, the physicians have been notified and in the -- have the option to dose reduce the patients that are enrolled. So although we do and have estimated a little bit of a delay as a result of this, I think we're still going to be substantially enrolled during the summer of 2022.

Maybe you could just touch on upgrade quickly. You mentioned UP-NEXT in your opening remarks and kind of what opportunity that opens up for you, but maybe you can talk a little bit about upgrading and how that sort of factors into your broader strategy?

So if you think about it, UPLIFT provides -- brings UpRi to the platinum-resistant patients, UP-NEXT brings us into the recurrent platinum-sensitive maintenance and then UPGRADE is really an umbrella study to understand the potential of combining UpRi with the standard of care. And we started -- we initiated in -- I think it was June or July of this year, the first combination, which is the combination with platinum. Platinum remains a very important therapy for ovarian cancer. As you know, platinum is used upfront, second line and third-line until patients become platinum resistant. Other agents have had challenges combining with platinum. We have the opportunity based on that differentiated tolerability profile without neutropenia and neuropathy to be able to replace [indiscernible] in the platinum combination, which would be very attractive for patients, given the profile of -- and that really tolerability issues with that combination. The study we are designing and we are -- in UPGRADE is that after a dose escalation portion, we will have an expansion portion. The patients will be given 6 cycles of the combination and then continuation with UpRi as a single agent. This would give us -- so you can think of it as induction followed by maintenance. This would give us very important information about the combinability and really set the stage as to where we want to go next with UpRi beyond UPLIFT and UP-NEXT.

Okay. Great. And I want to spend a little bit of time talking about lung cancer, and you're looking at 2 different assets there. One is UpRi and one is 1592. So maybe you could talk about the difference between those 2. I know, Tim, you touched a little bit on it initially, but maybe just what's the difference there? And how you're thinking about the strategy for those 2 assets?

So we have fully enrolled an expansion cohort of about 45 patients. We should be disclosing the data in Q4 for UpRi in lung. So what we have said and I can share with you is that we are finding that the prevalence of NaPi2b high is about half would decrease in ovarian. So we will be looking at the efficacy in the overall population and the efficacy in the selected population as well. As for 1592, we are in dose escalation, dose exploration, and we will have that data around year-end. The dose escalation, dose exploration is primarily in ovarian cancer because when patients -- because what we were looking for is to understand the similarities and differences between molecules and the selection of the Phase II dose. But those 2 data sets, the UpRi lung data set and the 1592 data set will be important information we will use to decide what is the future development of a NaPi2b ADC in lung cancer.

Can you just talk about the dosing you're using in UpRi in lung and how you're thinking about that now given your other sort of decision to?

So the expansion cohort in lung started at the same time as the expansion cohort in ovarian. So we started with 36 and then switched to 43. But I would say the majority of the patients -- the 45 patients I mentioned will be on 43.

Got you. Then just to clarify, it sounds like you'll later have data from both those assets and then figure out what you think the next step is. And could you walk, is it potentially going different directions in lung with both those or it select one or how you think about that?

I think one -- I mean, they were -- 1592 was intended to be a second shot on goal in lung. I think we will look at them. We'll look at the totality of the data. I think one thing we have talked about, and there's been a lot of work done here by Brian is to really think about what are, ahead of having the data in front of us, what should our go, no-go criteria be. Although we all think of lung as being a very large market opportunity, it's also a competitive market opportunity. And I think we've said for ourselves the go, no-go criteria ahead of the data. So we have some clarity on where we want to go next. And Brian can comment on what those benchmarks are.

Yes. Sure. I mean, obviously, lung is a very competitive area with a lot of activity in this space. And when we think about it, we think about the prevalence of the biomarker about a third instead of 2-thirds in ovarian. And we think about future-proofing the target product profile so that it's resilient in the face of future innovation in this space. If you look at the best analogs here, which are the randomized studies, they would put the performance of the standard of care somewhere between 14% and 24%. They happen to be a randomized study is happened to be from the pre checkpoint era. So if you look at post checkpoint data, that is primarily single center, single-arm and retrospective analyses that are a little bit harder to judge the certainty of. But suffice it to say, I think the bar is higher. Something that we would be excited about is 35% response rate. There's ample evidence that you can pursue that with a single-arm approach. World Lung was just this weekend. There are some other agents in this space that have put up 20% to 24% response rates. We'd be less excited about that because that would probably lend itself to combination development and add-on studies that are randomized with PFS endpoints. And we really want to make sure that we understood the profile before embarking on something like that. Again, to this point of future-proofing because you want to make sure that you're robust should the standard of care change in the next couple of years.

Got it. That's helpful. Maybe in the last minute here, you can just talk a little bit about the pipeline. Tim touched on some of this in -- early on, but maybe.

2 very exciting molecules moving into the clinic next year. We expect to be filing the IND in the earlier part of the year. One is B7-H4, a very interesting target, leveraging our Dolasynthen platform. And the other is 2056, our first Immunosynthen platform. And I would mention that Tim will be giving an oral presentation of the triple meeting on 2056. He'll be revealing the target and substantial preclinical data that really support the -- our decision to move this molecule forward and really our belief that it could have a very significant impact on patients. So look out for that update because it should be very exciting.

Okay. Great. It looks like we're out of time here. So why don't we end it there. Anna, Tim, Arvin and Brian, thanks a lot for your time today. We really appreciate it.

Thank you, Mike. Thank you for hosting us.