Following session is not open to the press.
Hello everyone and welcome to the annual Goldman Sachs Global Healthcare Conference. My name is Krishna Raghuram. I'll be moderating this session. It's my pleasure to introduce Barbara Weber, Chief Executive Officer of Tango Therapeutics. Barbara, welcome.
Thanks for having us.
Maybe for those who are less familiar with the company, could you start by perhaps providing a brief overview of Tango, your synthetic lethality discovery platform, how it's differentiated versus competitors in the market?
The basis for Tango really, as you pointed out, is synthetic lethality. The whole idea of that is that most of the drugs that have been developed in the last 15 years in oncology have been inhibitors that were targeting activated oncogenes. Every cancer has also associated with that tumor suppressor gene loss, and you can't directly target tumor suppressor genes because they're gone or inactivated. The whole idea was to be able to find a way to unveil a whole new category of targets against tumor suppressor genes to supplement all the work that had been done so successfully in activated oncogenes. This wasn't a new idea, but it wasn't really possible until CRISPR became widely used, and that was the basis for founding Tango.
Now that was in 2017, so fast forward now quite a few years, and we are now well into the clinic with our first target and have a pipeline coming behind that based on synthetic lethality. Just one interim point to make is that really the harbinger for the value of synthetic lethal targets is PARP inhibitors, and that was really the first discovery that was clinically relevant.
Great. Very, very helpful and informative. I guess, what are some of the current challenges in the synthetic lethality field and how can Tango address these challenges?
I do not think we all, really meaning us at Tango and the community as a whole, understood this as we started out, but the way synthetic lethal targets act is very different from activated oncogenes. I'd say first of all, there are a large number of synthetic lethal targets, but they are not all equally efficacious against a target. I think Chris Lord has talked about this as penetrance. That is sort of a version of a genetic concept, but it basically means how active in a cell is that inhibitor, even though it may be fully inhibiting the target. There is a range, whereas hitting an oncogene that a cell is essentially addicted to will essentially always give you a strong effect. Synthetic lethal targets can give you a range of effects depending on that target.
The other thing that we're seeing now specifically with PRMT5 inhibitors, and it's somewhat true for PARP inhibitors as well, although less dramatic, is that the really striking thing about that as a PRMT5 as a target is the durability as opposed to what people are used to, which are response rates. I think that's taken some time for us and other people to fully understand that this is what is special about these targets and will make a very important and unique contribution.
Got it. So maybe digging a little bit more into the pipeline and your lead program, can you give us an overview of TNG462 and then your next generation brain penetrant asset, TNG456?
Yes. TNG462 is our lead asset. That is a PRMT5 inhibitor, which is synthetic lethal with MTAP deletions. MTAP deletions are among the most common genetic alterations in cancer. In particular, lung and pancreatic cancer frequently have MTAP deletions. Our focus with 462, as we're nearing the end of our phase one two study, is development plans going forward for both pancreatic cancer and lung cancer, where we think PRMT5 inhibitors in general and 462 in particular will play a really important role going forward.
Great. Can you maybe help us understand the investigator enthusiasm here and how much data in terms of patient numbers, follow-up we might expect for the update second half of this year from the phase one two?
For our update later this year, what we plan to present, in addition to the full dose escalation expansion study cohorts and the data that support our selection of 250 milligrams a day as the dose to go forward with, we will also have a real focus on pancreatic and lung cancer. We will have more than 20 lung cancer patients with six months of follow-up and well more than 20 pancreatic cancer patients with more than six months of follow-up. We will, particularly in pancreatic cancer, also have a reasonable size cohort of second-line pancreatic cancer patients. It's those patients in particular that are really intended to support our plans for our first pivotal study, which will be in second-line pancreatic cancer. I think that those data will all be available with our update later this year.
What is getting investigators excited about this, I think, is the combination of the efficacy measured in particular by durability. I mean, it's pretty remarkable. We have a number of patients with difficult-to-treat tumors, including pancreatic cancer, that have been on for more than a year. I mean, I've been an oncologist for a long time. That's just not something that we've really seen before. The Bristol Myers Squibb data that was just presented at ASCO supports that as an effect of this target. On top of that, the patients tolerate the drug really well. And 462 in particular, we think, has a best-in-class tolerability profile.
Now you're talking about patients with esophageal cancer, pancreatic cancer, advanced difficult-to-treat cancers, and they not only are staying on drug for very much longer periods of time than standard of care, but they feel really good as opposed to the chemotherapy options, which they have that are not very effective and really not tolerable at all.
Great. You recently announced the dosing of your first patient for 456. Can you remind us of some of the preclinical data to date and which elements will ultimately correlate in your view to clinical success?
Yeah. You can sort of think of 456 as 462 with brain penetrance. The reason why that's particularly important to think about with MTAP-deleted cancers is because glioblastoma, almost 50% of glioblastoma have MTAP deletions. As you know, that is probably the single most difficult-to-treat solid tumor now that there's been advances in, and even now with pancreatic cancer and really moving into that much more treatable space. Making a PRMT5 inhibitor that can get enough brain exposure for those patients to have a chance at a meaningful response is important. That's what we think we might be able to do with 456.
Fantastic. Noting that this is somewhat of a successor to your previous efforts with 908, what learnings have you and Tango derived from that investigation that you hope to leverage and perhaps improve upon or implement with 456? What aspects or property of this molecule are perhaps differentiated from 908, which ultimately will drive your confidence going forward?
TNG456 is markedly more potent than TNG908, but the real difference that we think should matter in terms of glioblastoma is that we have tripled the selectivity for MTAP deletion. What that means is that we think we can deliver significantly more drug into the brain because of that. You can just give more drug so you can get more of it into the brain. TNG908 is a reasonably brain-penetrant molecule. The problem is you have to stop dosing as you hit toxicity. Because of the increased selectivity of TNG456, we expect that, waiting for the first PK of course, but PK-dependent, we expect that we will be able to dose TNG456 into the range where we are getting above the efficacy threshold in the brain. People say glioblastoma is a hard cancer. It is. I mean, there are some unique things about it.
It's a growing tumor inside a very enclosed space, but there's nothing magic about it. I think you get the right drug in there at the right exposure, and glioblastoma will respond like other solid tumors does.
Great. Now maybe taking a step backwards, noting that you have these two assets looking at the same target in your pipeline, help us frame and contextualize how you view the positioning of each molecule in the broader treatment paradigm and what is your criteria to perhaps call one of these programs?
We are very focused on moving 462 forward in all solid tumors with the exception of glioblastoma. It's, I think, a really excellent drug. The 456 program is really focused on glioblastoma because 462, we know, just will not get there. It's a big market in its own right. I think the other thing that's important is the flip side of that, which is that most brain-penetrant molecules will have some CNS toxicity. There are cancers like pancreatic cancer that virtually never metastasize to the brain. Having a really active and well-tolerated PRMT5 inhibitor that isn't brain-penetrant has significant value in all these tumors that are not likely to metastasize to the brain. Then having another one that is when you need it, I think, could be a very important strategic distinction.
Now you brought up lately Bristol's data. There are obviously some other PRMT5 MTA cooperative programs in the space. We've seen data from Bristol Myers Squibb, formerly Mirati, Amgen. There are others in the clinic from AstraZeneca. How do you view the competitive landscape here and maybe some of the differences between molecules? Given the clinical data that is out there and that the street, as we all know, is very likely to make cross-health comparisons, how are you talking and framing maybe expectations for your data from a response rate perspective, durability, safety, et cetera, compared to the others?
Yeah. And as you say, it's a complex field that's been generating a lot of interest for a while. There's a number of players out there. I think in the lead are Tango, Bristol Myers Squibb, and Amgen. All those molecules started in the clinic sort of in the initial wave. I would say there are qualities to the Amgen molecule that may hold it back. The public data, right, it has a slightly more difficult tolerability, safety profile, and the durability overall has been less than what Bristol Myers Squibb and we have already published on. We really think about Bristol Myers Squibb as our prime competitor, and it's a good molecule. We think that from the standpoint of the PK properties and the tolerability, 462 is better. I mean, just head-to-head data comparison.
We think in the end, because we can hit the target harder because of the PK properties and the tolerability, that will ultimately translate into at least the same, if not better efficacy. For us, the BMS data was really a very positive thing. Now, going back to your question, we'll come back to the BMS data, but going back to the competition, about a year behind that first wave is AstraZeneca. They haven't put out a lot of data, but presumably that's a good molecule. Then a year behind them are BeiGene and a number of other molecules. I think that it's on us, but we have a significant lead over all of those molecules.
Great. How are you approaching combinations? What have you seen so far preclinically, and what is your strategic development plan here with potential consideration for co-occurring mutations like EGFR or KRAS G12C?
Yeah, great question. Something we've been thinking about for a while, and our combination strategy is sort of divided into three camps. The first part are tolerability combinations that you have to think about for first-line studies if you have to combine with chemotherapy, knowing that you'll eventually have to combine with Pembro, those things. That is one bucket, and that is not a big priority right now for us because they are not on the critical path, and our real focus for time and resources is on 462 in pancreatic cancer. The second set of combinations that people have been talking about for a long time are the MAT2A inhibitor PRMT5 combinations. There is synergy, biologic synergy between those two.
The first study was the IDEAYA Amgen study, which never read out, but I think has been talked about in terms of why that was a difficult study to do because of the tolerability of that combination. There are others out there, and it's something that we think is a question that needs to be answered. Based on our preclinical data, we believe that a good PRMT5 inhibitor will always beat a MAT2A inhibitor single agent, and the combination may just give you a more complicated path forward. We're leaving that to other people to answer that question, but we're obviously interested in the outcome. The real focus of our combination strategy and what we think is going to be really exciting over the next year or two are co-occurring oncogenic mutations. The most exciting and important one right now are the RAS mutations.
The reason in part for that, it's not just the excitement around Revolution Medicines and the new class of RAS molecules. That's really important, and that's clearly central to pancreatic cancer and lung cancer development. Also, because the frequency of RAS mutations is such that essentially all MTAP-deleted pancreatic cancer is also RAS mutant. We are now just about to start a study where basically every MTAP-deleted patient with a RAS mutation, so more than 90% of them, are eligible to go on the study either in combination with daraxonrasib, the pan-RAS, or zoldonrasib, the G12D inhibitor from Revolution Medicines. Given the strength of the single agent data from all three of those molecules, and the really amazing preclinical data from that combination, we have high hopes for that combination. That's really game-changing for pancreatic and lung cancer patients.
Great. You've kind of touched on this a little bit throughout our conversation, focus on tolerability. Are there any overlapping toxicities with these combinations that you're watching out for? Do you think potentially dose lowering in either drug in these combinations would be advantageous?
All three molecules have really good safety profiles. I would say daraxonrasib maybe has a RAS issue and a little bit more GI toxicity, but overall very well-tolerated drug. We expect that the combinations will be very well tolerated. Now, we'll see, right? Things sometimes happen in combinations you don't expect, either good or bad, right? I think if you just look at the safety profiles of all the molecules involved and you think about the fact that they're completely non-overlapping pathways, I think they should be very well tolerated.
Great. Maybe taking a little bit of a pivot here, can you talk about TNG260, your CoREST inhibitor, which is being developed in STK11 mutation patients? Obviously, a large opportunity, which represents roughly 15% of non-small cell lung cancer. What's the status of the dose escalation, and when might we expect an initial update here?
Yeah, so this is a really, I think, innovative program that is sort of a riff on synthetic lethality, which is starting with the original data that we discovered in animal models and at the same time, almost at the same time, was published in retrospective analyses in patients that had been treated with, who had lung cancer treated with pembrolizumab, that STK11 mutations in tumors, which are inactivating mutations, really render cancers immune resistant. Those mice that have STK11 mutant cancers or the people that have STK11 mutant cancers really just are not sensitive to anti-PD-1 agents. What we did was an in vivo CRISPR experiment to try and find targets that reverse that setting of STK11 mutants and then made them be responsive to anti-PD-1.
As I said, it's sort of a riff on synthetic lethality to say, if you have an STK11 mutation, this will work. If you don't, you wouldn't expect to see any effect. That study now is in dose expansion. We have released the data on proof of mechanism, and we have shown that in patients, exactly what we predicted based on the animal models would happen, happens in patients. PD-1 expression goes way up, cytotoxic T cells come into the tumors. All the things that you would expect to have a tumor become sensitive to a checkpoint inhibitor all occur in those patient tumors. What will happen later this year is an update on that study looking at clinical proof of concept.
Great. Extremely helpful. How does 260 compare to other therapies that have been tried before, including LSD1 and pan-HDAC inhibitors?
They are all functionally related to each other because it's the CoREST complex that we think is driving this, and they are components of the CoREST complex. But 260, which we call a CoREST inhibitor because it does bind to a complex that contains HDAC, but it really only acts when that is part of the CoREST complex, is like 200 times more selective against this than an HDAC inhibitor. You know the pan-HDACs that have been in the clinic have been generally not well tolerated. They do not really have a therapeutic index. There is no patient selection strategy. It is really a pretty big difference between what has been tried before and TNG260.
You've shown some preclinical combination data here with an anti-PD-1. What's the path forward here as a combination? Could you give us a sense?
One thing that's really important is that the combination with PD-1 was engineered in from the beginning. Literally, the in vivo CRISPR experiments were done with a mouse anti-PD-1 to say, okay, if you knock out which genes in the presence of an anti-PD-1 will cause a response. What we found as one of the top targets was, in fact, the target of 260, which is the CoREST complex. Only in those STK11 deficient mice did we see that virtually all the mice got durable, complete responses, and they were essentially immunized against their own tumor by this effect. There is no path forward with 260 as a single agent. That was never an intent, and it isn't the biology, right? Even in the phase I study, every patient has been getting pembrolizumab in combination from the beginning.
Great. Now, perhaps more broadly, do you think that there's anything that the street is underappreciating with regard to Tango more broadly that we should be maybe paying more attention to on the forward balance of the year?
To be fair, I think we've probably been underestimated on pretty much every level, right? I get that. It's a tough time out there, but I think maybe the biggest things were really almost more on PRMT5 as a target. I don't think that investors have been really convinced until maybe pretty recently that PRMT5 was going to be an important target. If you don't believe our lead program is really going to be important, that's a big problem. I think there were other sort of rumblings around that. Maybe it's an important target, but Tango's way behind, or it's an important market, but Tango's not going to have the money to compete against Bristol Myers Squibb. I think none of those things turn out to be true, right? It is an important target.
I think it's going to be a cornerstone of pancreatic and lung cancer therapy going forward. I think we are absolutely not behind. I actually think we're in the lead for a couple of different reasons we could talk about. You know we have a lot of strong supporters, so I have no doubt we'll have the resources we need to do this.
Great. Really quickly, maybe touch upon capital allocation strategy for the balance of the year?
I really strongly focused on the pancreatic cancer development plan because our plan based on the data that we expect to present later in this year will be to start our first pivotal study next year. That will be a monotherapy study of 462 in second-line pancreatic cancer versus chemotherapy. That is our primary focus. We have made actually some changes at Tango to really make sure that our capital allocation is solely focused on that, well, largely focused on that going forward.
Great. Thank you so much, Barbara, for being here today. And to everyone else on the line, thank you so much. Have a wonderful rest of your day.
Thanks for having us.