So, I'd like to welcome Barbara Weber to the next Fireside Chat here, CEO of Tango Therapeutics. Barbara, welcome, and thanks for joining us.
Thank you.
So, we'll just jump right into the Q&A for the Fireside Chat, Barbara. So, we've seen now phase I data for PRMT5 inhibitors or for second-gen PRMT5 inhibitors from Mirati and Amgen last year, demonstrating single-agent activity in MTAP deletion tumors. Tango has two programs now in clinical development. Remind us of the preclinical profile of your agents and where you are in the clinic.
Yeah. Thank you. So, we have our first molecule, which is TNG908, that has been in the clinic for a while. That is the molecule that's brain-penetrant, and is in dose escalation, and as we've said, within the range that is clinically predicted by animal data to be efficacious. So, we can easily achieve those exposures with no DLT so far, and very good tolerability. TNG462 is our second molecule. It's not brain-penetrant, but it is considerably more potent and selective than TNG908. That molecule just started in the clinic last July, and the dose escalation for that molecule is going very, very rapidly.
I think it's a combination of investigator enthusiasm that's come as the POC data have come out from our competitors, and also the fact that because that molecule is very potent and selective, the starting dose is close to the active range, and we're able to dose double, so it's just going very, very quickly. The two are sort of almost caught up to each other in terms of being in the predicted efficacious dose ranges.
Okay. And what are sort of some of the preclinical features, or characteristics when you compare your inhibitors to the Amgen Mirati product candidates?
Yeah, I mean, I think, first of all, overall, they're more similar than they are different. They're all MTA-cooperative PRMT5 inhibitors. They're different scaffolds. Our two molecules are on a scaffold related to each other. The other two are completely different from ours and from each other. I think the potency and selectivity ranges across the four molecules, but interestingly, we have a slide in our deck that illustrates this probably better than I can describe it without a slide, but if you take the LU99 xenograft, which is the one. It's an MTAP-deleted RAS mutant lung cancer, a xenograft that all three companies have used and published data on. The activity of our molecules in that xenograft actually doesn't track with selectivity.
So, if you look at them across the range of the four molecules, TNG462 actually gives you the best response with a deep regression. That's 45x. TNG908 is sort of second with a minor regression. That's 15x. Mirati is third with stasis, and that is 75x. And then Amgen is sort of the least responsive, which is a tumor growth inhibition of about 70%, with 45x selectivity. So there's a lot of factors other than selectivity that go into sort of sorting through, and a lot of it's gonna come down to the clinical PK that is starting to emerge.
Okay, great. So, yeah, maybe just talking some more about TNG908, which is, I think, a more advanced program, and yeah, just remind us how far you are into the phase I at this point, and what investors should expect in terms of sort of near-term data disclosures?
Yeah. So, we have given guidance that we'll have clinical data on 908 this year. I think with Mirati and Amgen's proof-of-concept data out there, we need more than just to say, "This works." We're wanting to have a data set that will really help people understand what our next steps will be, how it might look compared to the other molecules. And hopefully, although we haven't given this specific guidance, is to say, with 462 catching up, that we may even be able to provide data from both molecules at once and answer questions about the bigger development questions of: How is your whole program gonna progress?
Right. And do you have some more color on the sort of patient makeup in the phase I study for 908? You know, is there a bias, for example, towards GBM, given that it's, it's brain-penetrant? You know, what should our expectations be there?
So, yes we've created a bias in a way with GBM. So, to start with, the 908 study has a whole range of different tumor types. I think when we provided the clinical PD update back in the spring, we had, you know, 16 patients and 12 different tumor types. I think we were just talking, Amgen released some more data last night, saying they've got 9 responses in 7 different tumor types. I think. So, we have a big range of tumor types in 908 for the dose escalation. We have been backfilling cohorts, though, to include GBM, and so we've been separate from the dose escalation cohorts. Those backfill cohorts have significant enrollment with GBM, so we create that sort of bias towards GBM there.
For 462, it's interestingly shifted a little bit, which is that we're starting to see more what we expected, which is that the tumor types that are going on study are more reflective of the population frequency. So, we're seeing lung, pancreas, cholangiocarcinoma, you know, the more standard common tumor types, as opposed to these rare tumors that went on earlier.
You know, GBM's obviously tough disease and, you know, just specifically in that, what are sort of efficacy benchmarks? So, you know, what types of, you know, activity, you know, can we, should we interpret there?
... Yeah, it is tough, and I have to say I've learned a lot about it in the last year or two. I think the patients that are going on to the 908 study with GBM are relapsed refractory patients that have failed the temozolomide, and often failed one or two other regimens, and those patients have a very short PFS without treatment. And I think what we're looking for there is a signal of activity in that tumor type, in that setting, relapsed refractory GBM, that gives us the confidence to move a little bit further or earlier into treatment. And I think there's a path forward to do that, that the investigators really support, so that we're looking at slightly less poor prognostic patients to really ask the question of what is the activity?
Right. And then, you know, again, stepping back, you know, you talked about 462 and the fact that you're seeing different sort of tumor types that sort of reflect the natural prevalence of MTAP deletion. And I think, Amgen and Mirati have sort of a more dedicated focus on non-small cell lung cancer, among other tumor types. You know, just if we step back, you know, how you think about potential histologist for you know, MTAP deletion or for PRMT5 inhibition? You know, is it, you know, based on our biology, would one expect activity to be similar, comparable across you know, independent of histology? Do you think there is some tumor types that are more responsive, perhaps, based on other, other markers?
Yeah, I have to say both the preclinical data and the emerging clinical data suggest there is no histology bias. Now, the real answer to that will come when all of us have filled expansion cohorts where we've got 20 or 30 patients each for each histology. But I have to say, the data that we have so far, both preclinically and clinically, suggests that there isn't a bias. Now, the reason to focus on things like lung and pancreas then becomes almost market-driven, right? Those are the biggest populations, and I think those are populations that none of us want to give away to our competitors.
But there's also the potential for, in my mind, clumping together some of the smaller indications and thinking about how one could get a histology agnostic approval for some subset of patients, that would be maybe those that their standard of care is chemotherapy, their response rates to chemotherapy are in the 20%-30%. You know, they're more homogeneous in terms of what you expect from the standard of care, even if the tissue type and the chemotherapy type are different. You could imagine designing a study that way.
Okay. And then, you know, as we talked about, there has been, you know, single agent activity, which is, which is great for a new target class. But, you know, Amgen is also working with IDEAYA, for example, combining their PRMT5 with, IDEAYA's MAT2A inhibitor. In that context, how do you think about developing TNG462, either as a monotherapy or as a combination?
So, there's two parts to my answer to that question. The first part is, I believe that all oncology needs to be combination. We've never made progress in any tumor strictly sticking with single agent. You get a signal there, and then you think about how to combine things to get deeper responses, more responses, longer PFS. I don't think these molecules are going to be any different with that, and we have some favorites that we're gonna start with. In terms of the MAT2A combination, however, the MAT2A PRMT5 combination is not an orthogonal mechanism to add together, hitting two different vulnerabilities in a cancer. It is, in fact, trying to get two molecules to get deeper inhibition of a single enzyme.
I think logically, if one can get full inhibition of that enzyme with a single agent, that's preferable because it makes subsequent combination studies better. You're not starting with a doublet, you're starting with a singlet. We feel the data support this strongly preclinically, that our molecules can get full inhibition of PRMT5, so are likely to not derive much benefit for combinations with the MAT2A inhibitor. I think that's probably true with Mirati also. On the other hand, the IDEAYA molecule, which is a MAT2A inhibitor, those molecules simply won't have enough single agent activity to move forward, so they'll benefit perhaps considerably with the addition of a PRMT5 inhibitor.
Great. Okay, great. Then, so, you know, again, looking forward to the sort of major update this year on, on TNG908 or perhaps both programs, it sounds like. And then, yeah, maybe switching gears to TNG260, your CoREST inhibitor, which is now in the clinic as well, targeting STK11-mutant tumors. Maybe just remind us of the mechanism and what makes this interesting?
Well, I think it's super interesting because it's such a novel approach to thinking about genetics and immuno-oncology approaches. But the idea behind this is that STK11 mutations create a tumor that is very resistant to being killed by the immune system, and therefore, resistant to checkpoint inhibitors. And what we did was identify in a mouse model that's STK11 mutant, what it is that you can target to reverse that immune evasion phenotype, and in this case, it's CoREST. There may actually be a few other things. There's some emerging data externally, but I think CoREST, for sure, is one of those things.
The idea is that you take patients with lung cancer, and there are others in addition to lung cancer, but about 60% of the lung cancer patients that otherwise don't respond to a checkpoint inhibitor and add TNG260 to the checkpoint inhibitor, and now you develop a responsive tumor. That's the question being asked in the clinical trial.
Great. And, you know, what makes it attractive? There's other mechanisms that are, you know, trying to pursue this, the PD-1 inhibitor-resistant market. You know, how does, you know, that mechanism, you know, differentiate from those?
... Well, what I think is the most interesting, of course, I'm biased, right? 'Cause I'm a geneticist, but it's strictly driven by the STK11 loss-of-function mutations. It's a specific mechanism that occurs when that molecule has been functionally deleted or inactivated, that upregulates cytokines and the cell surface receptors, or that would specifically predict that the cells are not able to respond-
Right
... to the immune system. And by adding TNG260, which is a transcriptional regulator, you reverse that. It's a very mechanistic approach, and it's totally based on the STK11 mutation. Doesn't seem to work in the absence of an STK11 mutation, which is good.
Yeah.
Right.
Yeah, great. And then can you walk us through the design of your study, your trial, that's ongoing now, and different considerations that were taken into account?
Yes, so thank you. It's a complicated trial because we tried to convince the FDA that we didn't need to do a single agent run-in for this study because there was no evidence that, that it would work, and we didn't expect any toxicity. That didn't go as we hoped, so we have a 21-day run-in with 260 alone, and then we re-baseline the patients and add the checkpoint inhibitor. That's the design. Everybody's getting the combination from the beginning, but in the dose escalation cohorts, everybody has to have a 21-day run-in with the single agent, 260.
Understood. And is it tumor agnostic, PD-1 pretreated patients?
It's tumor agnostic, but you have to have an STK11 mutation.
Right.
So. the non-lungs are usually breast, cervix, and a few... and pancreas. What was the second question you asked me?
Oh, are they PD-1-
Oh, PD-1-
Pretreated or not?
Pretreated. Most of them are, because most of the lung cancer patients will have gotten that as standard of care.
Yeah.
The other ones, not necessarily. So cervical, breast, those patients may not have seen checkpoint inhibitor before.
Gotcha. Understood. And how has enrollment been progressing so far, and when might we be able to get a first disclosure?
I think it's been going well, and on top of that, we got another little added bonus, which is the PK for the molecule turned out to be excellent, so we're closer to where we thought we would need to be for the efficacious dose range. So, we're not giving specific guidance, but there's a possibility we might have some data later this year.
Okay, great. And yeah, and-
Kwok Wong has a very interesting poster at AACR with some additional clinical data, preclinical data on this.
Gotcha. Super, so we'll keep an eye out for that.
Yep.
Yeah, and then, as we look through, or as you look at the sort of phase 1 data and, you know, especially close to RP2D, how do you think about the hurdle, you know, efficacy hurdles that you would consider interesting to move this forward into-
TNG260?
Yeah.
Yeah, I think the bar is pretty low here because the, these are lung cancer patients that, in general, just don't respond. So, the PFS is in the sort of 2-3-month range, and the ORR is, you know, 5% or something like that. So, historically speaking, the bar is low, so I think we'll, we'll have to see, but I think...
Right. Okay, great. So maybe switching gear to your next program, TNG348, which is your USP1 inhibitor, which is quite interesting. I think there's a number of other USP1 inhibitors in development as well. Maybe just talk about the features of your molecule and possible differentiation from others.
So, I have to be upfront about that and say all of the USP1 inhibitors are sort of more similar again than they are different, and it's a particular issue with this class of molecules because it's been difficult to derive totally novel chemicals matter around the target for any company. So, everyone's got their own scaffolds for IP purposes, but they have a pretty similar mechanism of action against an allosteric pocket that's similar in all cases. So again, it'll come down to sort of PK and development strategies to really differentiate them. I think on the plus side, for us, KSQ, which is the now Roche molecule, we've never seen the data, but if you use the Roche co-development or licensing deal as a sort of surrogate marker for activity, we feel confident that that's been de-risked by that-
Right
... deal.
Just stepping back again, just remind people of USP1, where's the opportunity, and where does inhibiting USP1 make sense from a histology perspective?
Yeah, so USP1, like PARP, is synthetic lethal for BRCA1, BRCA2, and other homologous recombination repair deficiencies. It also. We've also shown that in some mechanisms of PARP resistance, it is still active, and it's still synergistic with PARP in the presence of PARP resistance. So, I think the market is basically the PARP market, including the PARP resistance market-
Right
... which is significant.
Is your phase 1 study enriching for BRCA mutations? How should we think about the trial?
Yes, with the... Which is an interesting point. The KSQ dose escalation was in all comers, not selected for BRCA1 or BRCA2 mutations. Our dose escalation and expansion requires either a BRCA1 or BRCA2 mutation or a documented mutation in one of the other HRD repair deficiency genes.
Okay, great. Any comments on how the study's been going so far, and you know, whether you might be able to share some data this year?
I'm less optimistic that we'll have data on that this year. We just dosed the first patient-
Right
... in December, and we still are in the process of getting most of the sites up and running but shout out to Tim Yap and MD Anderson for their enthusiasm in getting some patients on.
Sounds good. Anything else we should watch out for USP1, any on target tox concerns or anything that we've learned from the competitors?
Well, the GLP tox study is that it's a very clean molecule. We see very little. I mean, we get, like, some knee swelling in the rats at the highest doses, and I know that the KSQ dose escalation, the press release, says that they never reached MTD. They, I don't think, had DLTs. They just decided to go into the combination because they were tired of dose escalating-
Mm-hmm
... which doesn't surprise me, right? It's a sort of a 200x selectivity, so.
Right. And so, okay, so that's all I had on USP1. So you guys are obviously one of the companies that has a, you know, very innovative pipeline through a product engine, and I'm sure you're doing more beyond those four into the discovery. How do you think about leveraging that capability for strategic partnerships down the road? Is that something that you're considering? And what are, what are areas of, you know, discovery activities that you have?
Well, not surprisingly, probably, you know, given our strong investment in the PRMT5 program, what we've been interested in thinking about developing novel targets that could be interesting combination partners. So there's some of that that's happening. There's some novel synthetic lethality that's in there, and, you know, there may be some interesting additional PRMT5 molecules coming, so.
How very interesting.
We are active on our... We're keeping our target discovery and drug discovery pipeline as active as it was when we built the company.
Great. And then maybe just to wrap up, just to remind folks of your cash runway and, and you know, anything that you should know there.
Yes, we are in a very fortunate position from that standpoint. We have cash now into late 2026. Our balance as of 12/31, unaudited, was around $230 million, and we raised $42 million with our ATM just after J.P. Morgan, so our pro forma balance now is around $370.
Great. All right.
Which will get us through all of these readouts.
Sounds good. So looking forward to hopefully seeing updates on four clinical programs this year, maybe three at least, it sounds like. And, yeah, Barbara, I'd really like to thank you for the fireside chat, and I think we can conclude here. Thank you.
Thank you. Thank you very much.