Tango Therapeutics, Inc. (TNGX) Earnings Call Transcript & Summary

Okay. Perfect. Thank you so much. My name is Pete Lawson, and good morning, everybody. Welcome to Barclays Healthcare Conference in Miami, and really pleased to have with us up on stage management from Tango, Barbara Weber, CEO.

And the first question, I guess, main question has been over the last year or so, which is around PRMT5. Kind of what data do you expect to show this year? And are there any particular trigger points to showing that data set?

Yes. I think -- well, I think after a long way, we will have clinical data on TNG908, our first PRMT5 inhibitor this year. And we will have a substantial number of patients, although caveating that with one of the primary reasons that it's been a while is that we were mandated by the FDA with a low starting dose and 50% dose increments rather than the normal dose doubling. So there will be a number of patients, but a number also at inactive doses. And you may recall back in May of 2023 that we had said we were in cohorts 4 and 5 and still not to expect activity based on preclinical projections. So having said that, a number of patients that we expect to be able to report on this year. And we will have, I believe, enough GBM patients included in that to answer the question whether or not 908 has activity in GBM.

Remind me again, the 50% jump in dosing, why was that?

It was -- it happened because in rodents, there was some nonlinearity in exposure that had the exposure going up faster than predicted with dose. So we were asked to be cautious with dose escalation. That did not happen in patients, in humans. The PK is very linear and very predictable, but it did lead to quite a long dose escalation.

Got you. And the -- where are you for dose level now? Or what are you enrolling in?

We haven't specifically given a number, but we are well into the exposure ranges predicted to be active by the preclinical models.

Okay. What were those ranges where you thought you were going to see activity?

In the -- you mean in terms of dose or...

[indiscernible] 5 sits kind of thing or...

That's what I'm saying. We haven't given that exact number, but it's in a range that's less than a gram dose. I guess I could say that.

Okay. Perfect. And kind of just in the context of everybody else's data, kind of what are the puts and takes that we should be thinking about?

Yes. And I think that's also an issue for us to think about in terms of clinical data updates, which is that Mirati and Amgen did provide the POC data last year. I think they really showed that these MTA cooperative, PRMT5 inhibitors will be active. So I think it will be important for us to be able to benchmark against them and give people some idea of where our molecule stands in comparison to the data that are out there.

Got you. And in particular, around dosing, is there differences that could drive differentiation or benefit or...

I don't know if that will be around dosing or not. What I would say is that there's some interesting and not yet fully explained data from both Amgen and Mirati, now BMS. I think if you look at the Mirati data, they had activity in the range of 1 out of 3 patients per cohort, so people are roughly estimating 30% ORR. I think it's a little tricky in a Phase I dose escalation because of the small numbers at each dose, but let's just go with that. And say that they had sort of similar activity and toxicity over a wide range. So I think they'll probably have to do some more work to understand what their optimal Phase II dose will be, but they're certainly in a range that seems to be tolerable and active. I think the Amgen data, they, I think, may have a bit of a bioavailability issue, and they got up to pretty high doses with perhaps some nonspecific toxicity. So we'll have to see what happens with them. I can say that our molecules, 908 and 462, are very well tolerated at the doses that we're currently giving, which, as I said, is well within the range of what we would expect to be active based on the mice, and for both toxicity and tolerability looking really good and not yet having reached an MTD.

The bioavailability for the Amgen molecule, what were the hints there in that data set?

Well, I think if we look at our data, the hints are that there are -- we get about 3x as much exposure per dose on a gram-per-gram basis. So there's that. I think they ended up also with a fairly potent selective molecule having to go quite high in terms of dosing. But they are dosing once a day. And I think they had perhaps hinted out and would make sense that considering a BID regimen might be helpful to that.

Got you. Okay. And the next updates for Amgen, Mirati, what should we be honing on to triangulate to yours?

Yes. Well, again, I think it's just a matter of looking to see what everybody's molecules look like. I think the big themes are, though, the similarities are likely to be more than the differences. The one important piece of this is that the preclinical data that we have, that others have and in the clinical data that are out there suggest that histology is not a predictor of sensitivity, that MTAP deletion is, in fact, the primary predictor. So in suggesting that as we hoped and as the preclinical data supported, these molecules will be active in a range of histologies.

Got you. The next-generation molecule, so...

462.

462. What are the design features there that could make it best in class?

Yes. I think 462 is a really nice molecule comparing it to 908 and it's more potent, it's more selective against MTAP-deleted cells. It has a really nice PK profile with a very long half-life and a nice flat curve that keeps the enzyme inhibited very nicely over a 24-hour period without dips below trough. So I think those things are all really positive. It appears to have a really long off rate, maybe even longer than Mirati's based on terminal PD studies. So one downside compared to 908 is it is not blood-brain barrier penetrant. So GBM has been excluded from the 462 studies, and that's something where we would need to focus with 908.

Got you. The -- and so it kind of leads into indications where you had preferentially select 908 versus 462.

And that would be GBM, and perhaps even consideration of things like non-small cell lung cancer, where the propensity for brain metastases is high.

Interesting. Okay. And then -- and are there any preferred routes for 462 or it's everything?

I think it could well be that it's everything, but GBM. We have to see what the data look like. But I think we've never had the intention and you and I have talked about this going forward in the same indications with both molecules. So I think we would focus -- assuming the data support it, we would focus on GBM with 908 and other solid tumors with 462 because even if activity is equivalent between the 2, then there's still this PK advantage that's potentially more positive for 462.

Okay. And is it a case that 462 moves where there's white space versus the competitors or do you...

You know, that's probably a strategy that would have worked potentially before the IRA. I think the focus now is on the big indication so that that's something that can be addressed over the long term. And so I think all of us are focusing on pancreatic cancer and non-small cell lung cancer. But thinking how to do that versus the competition is something we'll have to focus on, and we'll see what the data look like.

Got you. You wouldn't think of developing another molecule for different indications as well to...

I think that's a partial strategy to think about, right, in the [ IRA ]. So yes, that's something we talked about.

Okay. Is that something in the works or...

Well, we haven't disclosed any new pipeline information recently, but certainly the idea of a molecule that is potent and selective as 462, but brain penetrant would be something that would be appealing to have.

Got you. Instead, I guess from the preclinical data that we've seen so far from yourself, Amgen and Mirati, are there other indications we should be thinking about that beyond that kind of IRA focus?

I don't think so. I mean I think there was some discussion, for example, about mesothelioma because some of the preclinical lines are very sensitive and there were a few responses with Mirati. I think that, that really more has to do with the patients in the study than the fact that any indication is more sensitive to others than others to these molecules. So again, it will have to be driven by the data.

Got you. And then just in light of Project Optimus, do you take 2 doses forward? Or how should we think about that?

I mean, I think my -- the ideal scenario for us would be to have enough backfill data from the dose escalation that we can support a single dose into escalation to the FDA. And that's one thing that we're working on, backfilling cohorts of interest in the dose escalation phase. Some companies have, in fact, just gone ahead and put 20 patients in on a second dose. I'm not sure that, that makes any logical sense, but it is something that's sometimes been requested. So we'll just have to see how that goes. And I would say, again, our molecules are very well tolerated and I would expect, particularly 462, that's very selective for MTAP-deleted cells. There may be a pretty broad range where it's active and tolerable at multiple doses, and we'll just have to see what that looks like.

And then Amgen ideas approach with this MTA -- sorry, [ MAT2 ] inhibitor. Is that something you imitate? Or is there another way around that? Or is it just your idea of hitting this kind of axe is hard?

Well, I think that MAT2A combination is something we've spent a lot of time talking about and thinking about for the last 3 or 4 years, but I think what we believe, based on the preclinical data and what we're seeing now, I think, particularly with 462 supports this, which is that what MAT2A inhibitors do is decrease SAM levels, and they decreased SAM levels equally in normal cells and in MTAP-deleted cells. So if you sort of think about it, I guess graphically, here are SAM levels, which, if you will, are sort of is an activating co-factor for PRMT5. And in normal cells, there's really very little MTA. The MTA-cooperative inhibitors, which are what we're all working on now, really only work when there's high levels of MTA and that happens when there's an MTAP deletion. So in those cells, this ratio suddenly switches, right? So MTA is up here, SAM is down here. And what an MTA inhibitor does is it modulates the SAM levels. So if there's already a really big difference between MT and SAM, this doesn't do much in terms of inhibiting the enzyme. So our feeling is that in that setting, and our data supports this, if you can fully inhibit PRMT5 with doses that are very tolerable, there's no advantage to adding to an MAT2A inhibitor, there's really only a potential downside of both complexity for future combinations and toxicity. So we already are aware, and as I think I mentioned, we are in active dose ranges predicted by preclinical models with both 908 and 462, meaning we already know that we can give doses that are -- give us exposures that are as high or higher than where the mice get optimal responses with really good tolerability. So I don't see any advantage to us to moving forward with the MAT2A inhibitor combination. Now obviously, we'll look at the data, and if the data support it, it's something we can certainly do. But I think if you have an MTA-cooperative PRMT5 inhibitor that can fully inhibit the enzyme alone, you won't get any additional benefit out of that MAT2A combination.

What are the future combinations, you think?

Well, some of the ones that we're most interested in are really driven by the genetic abnormalities that occur in MTAP-deleted tumors, I think are sort of our -- our priority combination is a CDK4/6 inhibitor because every tumor that has an MTAP deletion also has a CDKN2A deletion. And that combination with the CDK4/6 inhibitor because of that biology is very synergistic in preclinical models. So that's one that's very interesting to us. And then I think thinking about other oncogenic drivers, probably RAS combinations are the next most interesting to us, particularly, I think -- I mean, lung is an obvious one with the G12C, but pancreas, the PANRES -- the RevMed-PANRES molecule and the G12D molecules that are coming could be really important combinations in big indications of lung and pancreas.

Got you. And is there a particular CDK4/6 you'd go with? Or is it just kind of standard of care?

I think that any CDK4/6 inhibitor theoretically could work in combination. The advantage of abemaciclib is that it's brain penetrant. So you might think about that in combination with 908, where you're looking for brain penetrant activity.

Yes. And the [ pan-KRAS ] assets, the RevMed, you're on the Board, is that's something that...

Definitely of interest.

Interesting. Exactly. Okay. So kind of watch this space for...

Watch this space.

Perfect. And I guess beyond that, are there -- with those kind of combinations, does that change the dose that you'd go in with a combination? Or do you keep the...

It's an interesting question. I think if you are any situation where you have synergy or you have a drug interaction, you have -- you might need to decrease the dose of a molecule, but you get the same exposure. I think for us, again, we want to go in with the lowest dose that has full activity, and we have good tolerability, particularly with 462, we have little risk of drug-drug interaction. So I think for us, the likelihood we would need to dose reduce, particularly with 462, is low. But again, that's why we always have to do dose finding before we do a full-on combination study.

And then I'd love to move on to the CoREST inhibitor. When do we see data on there? Is that a '24 event?

So we haven't yet given guidance on our CoREST inhibitor, TNG260. I think that's a really interesting program, which, as you know, is designed to reverse the immune evasion that's caused by STK11, loss of function mutations. That trial started in the summer of last year 2023, and it's actually going really well. We also had a nice surprise with the PK on that molecule being significantly higher than predicted by the animal models. So I think with that, we're in or pretty close to the predicted active dose range based on the preclinical modeling. And while we haven't given guidance, I think that program will read out maybe more quickly than we expected. So maybe this year, maybe next year, we don't know yet, but it's pretty interesting, and it's going pretty fast.

Got you. And so going really well is enrolling really fast?

Enrolling well. I would say it's not -- the PRMT5 programs are easy to enroll, long waiting, less -- lots of enthusiasm. This is a complex study, but I think it's enrolling without any difficulty in moving actually more quickly than we expected.

Got you. And that's because it's easier to identify the mutation or...

People do often have pre-identification of the patients with the STK11 mutations. It's on all of the commercial panels. It's on most of the academic panels. And I think it's -- 2/3 of it is in lung cancer. So people are quite used to screening lung cancer tumors for actionable mutation. So yes, that's actually gone pretty well.

Okay. And the trigger for that initial data, is there responses? Is it getting into high doses?

No, it will be responses. I think what we know is, of course, going to have to be based on historical data, but there are a lot of data showing that patients whose tumors have STK11 loss of function mutations just really have primary resistance to pembrolizumab. So in that setting, the response rates are low, 10% or less, and the PFS is low, usually 4 months or less. And I think it's those historical benchmarks that we're going to be looking forward to seeing whether there's evidence that TNG260 can meet or beat those historical benchmarks.

And what would you want to show in the initial data set?

I think we'd want to show that, in fact, the activity of that molecule gives us response data, ORR and durability data that are better than what's historically been seen and studied in the STK11 mutant lung cancer patients.

Okay. And how much data would we see?

How much data sort of depends on where and when we see that activity, right? We are -- we will backfill cohorts as well. We will have both lung and non-lung cancer patients. And when we feel we have a data set that really can address those points we were just talking about is when we'll release the data.

And it would be a case of like is one response good enough? Or is that sometimes misleading...

I don't think one response is good enough. One response can always happen. I sometimes say 3. But I think it just depends on the context and what the denominator is.

Got you. Okay. And you clearly...

Don't forget this is in combination with pembrolizumab. So that's a complicating factor that we need to address to make sure that people feel confident, we're just not seeing pembro responses. Now so far, I believe all the lung cancer patients on this study have failed pembrolizumab. So that's also helpful in addressing the question of, is this a TNG260 effect? Or is this just a pembro effect?

Okay. So you kind of probably err on the side of more patients business...

I think so. Yes. And there's skepticism. No, I'm a geneticist, I'm skeptical of immunology, too. So just want to make sure that we understand the data before we put it out there.

Did -- so PD-1 is one of the combinations? Are there other combinations you want to explore or...

Well, I think for right now, the first -- the focus really is on understanding whether TNG260 does reverse the checkpoint inhibitor resistance that is caused by STK11 mutations. Once that happens, then probably the next most obvious thing is thinking about combining it with a standard chemotherapy regimen because those -- that's what allows you to get to first line in those lung cancer patients. I think for the other 1/3 of STK11 mutant patients tumors that are not lung, there's a couple of different things to have to be evaluated. But there's opportunity there in multiple others as well.

Got you. So the kind of the not lung would be -- you'd think about standard of care as a combination...

But again, it's complicated. So the non-lung is breast, cervical and carcinoma of unknown primary, which is usually breast or lung. So I think those are areas where we'd have to see -- you'd have to understand what's the response rate that we saw theoretically with the combination of 260 plus pembro? And what does that look like in terms of standard of care?

Got you. And I'd like to move on to the USP1 inhibitor in the last few minutes. Just, I guess, thinking about KSQ, what license, Roche, kind of just how you view your molecule differentiating versus KSQ?

Yes, I think that -- and you'll see this at AACR, there's now 7 USP1 inhibitors out there with the AACR abstracts. All of the -- the only way we and others have found to inhibit USP1 is through a single [ allosteric ] site. So you might imagine that there's a fair amount of similarity in the molecules that have that mechanism of [ action ]. So I think we just have to see what the clinical data looks like because even as we were talking about with PRMT5, the PK ends up being an important component of that.

Okay. Is it kind of wait to AACR to really kind of understand if there's differentiation emerging for these molecules or...

We'll have a look. I have to say, based on what we already know from the AACR abstract, there's no molecule that's out there that seems fundamentally different than the original former molecule, the KSQ molecule, ours. I mean, they're all protected by the IP around them, but the actual mechanistic approach of an [ allosteric ] inhibitor to this specific site, they're all taking -- they all have that same mechanism into that same site.

Okay. And just same question, is there a trigger for Phase I data to be released?

Well, for us, that trial just started, and we just dosed the first patient in December, and it's both a single agent and combination trial. So we're currently dose escalating with single-agent TNG348. We will, as we get to pharmacologically active dose ranges, do a parallel dose escalation in combination with the PARP inhibitor. So I think I would be surprised if we haven't given guidance, but I'd be surprised if there were data this year.

Got it. And it seems what KSQ started in December '22. So do you feel like you're a year or so behind or...

I think we're about a year behind based on a couple of different things. One is that they did sequential dose escalation, single agent and then PARP. They also did dose escalation without any selection for BRCA1/2 mutant or HRD mutant tumor. So they got very little information about activity, dose responses in that setting. They started their PARP combination, I think, in January or February of 2023. And the Roche license deal happened not very long after that, maybe 4, 5, 6 months after that. We do know that in June of '23, based on clinical trials, they halted that trial, right after the Roche licensing announcement, and that trial was stopped for about 6 months probably because of supply issues. We don't know for sure, but that's, I think, a reasonably educated guess. So between the fact that we are doing a parallel combination with PARP and single agent, and this 6 months gave us a little bit of extra time to catch up because we -- I'll tell you what, we think that the supply issue was a similar problem we had, but we solved it before we started in the clinic. So all of those things, I think, get us to about the year behind.

Got you. That's a play issue relates to like the API or something or...

It does. It has to do with the hydration form of the molecule that can be a bit tricky. So you have to move to solve that problem. You have to move to a spray dispersion formulation.

Anything most will follow those...

I think there's a reasonable likelihood that all of the USP1 inhibitors will have that problem based on what I told you before that there's some similarity of what you need to have to get to that [ allosteric ] pocket.

Interesting. Okay. And well, thank you so much. Wrap it up there. Always a pleasure speaking.

Thank you. Great. Thanks, you too.