WEBVTT
00:00:17.120 --> 00:00:21.440
Hello and welcome back to Better BioForma, the official podcast of Bioprocess Online.
00:00:21.760 --> 00:00:32.719
I'm your host, Tyler Menechello, and on this episode, I'm joined by Johanna Kaufman, Chief Scientific Officer at Deck Bio, a company developing multipeptide MHC targeted T cell engagers for solid tumors.
00:00:32.880 --> 00:00:34.960
Johanna, thank you so much for joining me today.
00:00:35.520 --> 00:00:36.399
Thank you for having me.
00:00:36.479 --> 00:00:39.520
Thank you for the invitation, Tyler, and I look forward to the conversation.
00:00:39.920 --> 00:00:40.640
Likewise.
00:00:40.799 --> 00:00:52.880
So to start, why don't you tell our audience a bit about Deck Bio and what the heck I mean when I say multipeptide MHC-targeted T cell engagers, which I don't think I could say five times fast, even if I if I I know it's a mouthful.
00:00:54.079 --> 00:00:54.399
Yes.
00:00:54.560 --> 00:01:00.079
So DeckBio is a pre-seed stage startup company based in Cambridge, Massachusetts.
00:01:00.320 --> 00:01:05.599
And we are, like you said, developing novel T cell engagers for solid tumors.
00:01:05.840 --> 00:01:22.959
T cell engagers are an up-and-coming modality in the immunooncology field, and they're essentially cross-linking and affector immune cell T cells specifically to cancer cells, and then mediating essentially cell killing.
00:01:23.680 --> 00:01:39.519
This modality is well established in hematological malignancies or blood cancers, but translation into the solid tumor space, which is actually 95% of all cancers, has been lagging behind and is so far restricted to very niche indications.
00:01:39.680 --> 00:02:00.799
And it is our mission at Dech Bio to take the learnings from the clinical and design perspectives of T cell engagers together with the biology in solid tumors and address this unmet need and therefore bringing T cell engagers to broader patient populations with solid tumors specifically.
00:02:01.120 --> 00:02:08.479
We're doing this in the way of choosing peptide MHCs as a target class.
00:02:08.719 --> 00:02:18.000
The reason behind this is that traditional cell surface targets often have residual expression on healthy cells.
00:02:18.159 --> 00:02:32.159
And while that may be okay for other modalities like antibody drug conjugates, for example, as long as you have the right dosing window, for very potent modalities like T cell engagers, that is actually a safety liability.
00:02:32.400 --> 00:02:41.439
So we want to choose targets that mark the cancer cells specifically that are expressed as exclusively in tumors as possible.
00:02:41.680 --> 00:02:52.319
Now, the problem is that most of those targets are actually intracellularly expressed and are not naturally easily accessible to biologics like T cell engagers.
00:02:52.479 --> 00:02:59.599
And to overcome this hurdle, we are targeting these molecules through peptide MHC complexes.
00:02:59.840 --> 00:03:14.879
These are generated as part of natural immune surveillance mechanisms of the body, where essentially every intracellular protein gets chopped up into peptides and sampled on the cell surface and displayed in the form of these peptide MHC complexes.
00:03:14.960 --> 00:03:18.879
And that's what we're using as a target class, as the foundation.
00:03:19.039 --> 00:03:22.800
And then we're taking this approach to the next level through multi-targeting.
00:03:22.960 --> 00:03:30.479
And I'm sure we'll get into the rationale and the ways we do this in a little bit more detail as part of this conversation.
00:03:30.719 --> 00:03:45.120
We are currently focusing our efforts on our lead asset, DBX01, which recognizes a signature displayed in the most prevalent HLA allele of the Western world, HLA A0201.
00:03:45.520 --> 00:04:00.800
And we are driving the RD process towards entering the clinic in 2028 with additional pipeline programs that leverage the same multi-peptidemia C-targeted concept in the context of additional HLA alleles.
00:04:02.000 --> 00:04:02.800
Great, Johanna.
00:04:02.960 --> 00:04:07.439
Thank you so much for that succinct and uh comprehensive overview there.
00:04:07.599 --> 00:04:08.560
I appreciate that.
00:04:08.719 --> 00:04:18.000
And you're right, we will be getting into uh the rationale behind multi-targeting and kind of the structural um molecular design underpinning that.
00:04:18.160 --> 00:04:20.240
So I'd love to hear from you, you know.
00:04:20.319 --> 00:04:30.560
Yeah, of course you hear about the challenges of multi-targeting, but I before we get into that, I want to hear from you about the rationale to choose multi-targeting.
00:04:30.639 --> 00:04:33.759
Why do you think that that's necessary from a therapeutic standpoint?
00:04:33.839 --> 00:04:43.920
And is it something you mentioned that uh this we have T cell engagers in the blood cancer space, not not really much in the solid tumor space.
00:04:44.560 --> 00:04:56.319
I've heard from others that because of the the makeup of antigens that's associated with solid tumors, that uh it kind of lends itself to to multi-targeting or more or less demands it.
00:04:56.399 --> 00:05:01.279
So yeah, I'd be curious to hear um why you think it's necessary from a therapeutic standpoint.
00:05:01.839 --> 00:05:09.519
Yeah, I think the main reason why multi-targeting is necessary is heterogeneity.
00:05:10.000 --> 00:05:29.519
You know, and when we draw up mechanism of action for a drug, we generally depict the tumor as a homogeneous, uniform mass where every cell within that tumor, every patient seemingly expresses a particular target or has a particular aberration that a drug can address.
00:05:29.680 --> 00:05:37.839
And then therefore, theoretically, every one of these cancer cells or every patient in a cohort will respond to therapy and be cured.
00:05:38.079 --> 00:05:41.600
The reality is that that's not how this really works.
00:05:41.759 --> 00:05:46.560
We actually have quite uh heterogeneous tumors with different parameters.
00:05:46.720 --> 00:05:54.319
And you can think about this at the level of different individuals, different patients within the same tumor type.
00:05:54.639 --> 00:06:04.639
So if we both were in the unfortunate situation to have, let's say, non-small cell lung cancer, your molecular makeup might not be identical to mine.
00:06:04.800 --> 00:06:13.439
It might also not be that every lesion within a patient's body has the same makeup, or every cell, even within a tumor lesion.
00:06:13.600 --> 00:06:19.759
And there are different types of targets that might be expressed, or at least at different levels.
00:06:20.160 --> 00:06:27.600
This phenomenon of heterogeneity is much more pronounced in solid tumors than it is in blood cancers.
00:06:27.759 --> 00:06:36.639
And so you're right, this problem is much more pronounced and a lot much larger hurdle that needs to be addressed in solid tumors.
00:06:36.959 --> 00:06:55.680
So if we're focusing on one aspect of this heterogeneity, which is target expression, we often realize that target is expressed at different levels in different cells or patients, or only a fraction of the cells actually expresses or displays a target.
00:06:55.839 --> 00:07:11.680
So if I make a T cell engager against a single target, the traditional bispecific T cell engager, I might only eliminate a subset of the cells, the remainder grows out, and I ultimately have not achieved a desirable outcome for that patient.
00:07:12.000 --> 00:07:24.959
By going after multiple targets at the same time, I have a higher chance that I can cover the different types of subsets of cells within a tumor nodule or different lesions or different patients.
00:07:25.120 --> 00:07:35.279
And therefore, I can increase the likelihood of response, I can increase the number of patients that might respond to a therapy.
00:07:35.519 --> 00:07:40.879
Um, and that is obviously beneficial from an outcome, clinical outcome as well as commercial perspective.
00:07:41.040 --> 00:07:52.959
And that is why we strongly believe that multi-targeting in general, whether you do this through peptide MHC complexes or otherwise, and whether you do it in our unique way or otherwise.
00:07:53.360 --> 00:08:01.120
But the concept of multi-targeting, we believe, is extremely important for durable responses and broad patient populations.
00:08:01.759 --> 00:08:02.160
Yeah.
00:08:02.639 --> 00:08:04.000
Thank you, Johanna.
00:08:04.480 --> 00:08:13.759
And I was particularly fascinated by um your ability to target a single binding domain with your T cell engagers.
00:08:13.839 --> 00:08:21.680
And I'm curious if you could explain how you are able to do that to uh target multiple peptide MHC complexes with a single binding domain.
00:08:21.759 --> 00:08:25.839
I don't think that's common in the T cell engager kind of landscape.
00:08:26.079 --> 00:08:28.480
Correct me if I'm wrong, but that is quite new.
00:08:29.040 --> 00:08:32.080
That's accurate, at least for multi-targeted uh T cell engagers.
00:08:32.240 --> 00:08:37.679
So our molecule is uh in some ways a very simple bi-specific molecule.
00:08:37.840 --> 00:08:42.000
One arm obviously targets the T cell side of things, CD3 in our case.
00:08:42.159 --> 00:08:44.000
We're not reinventing the wheel here.
00:08:44.159 --> 00:08:54.399
The unique aspect is the cancer-engaging arm, which is a single binding moiety that recognizes multiple targets.
00:08:54.720 --> 00:09:29.360
The reason why this is possible is that during target identification, we are identifying signatures of multiple target proteins that are expressed specifically in cancer, not in healthy cells, that derive peptides that are either identical to each other or highly similar to each other, such that these peptides get loaded into the same HLA allele and can therefore then be recognized by a single binder.
00:09:29.759 --> 00:09:46.720
The key here is that these different peptides, as they lay in the peptide MHC binding groove, display the same or almost the same surface topology or 3D epitope, if you will, that a binder can recognize.
00:09:47.039 --> 00:10:02.720
Any sequence differences of these peptides are pointing inwards into the peptide MHD binding groove inside into the complex, such that they do not contribute to that surface topology.
00:10:02.799 --> 00:10:18.799
So, from the perspective of the binder, it actually looks like the same thing, except for that it is actually in the case of our lead acid DBX01, two peptides that get loaded into the same HLA allele, again in our case, AO201.
00:10:18.960 --> 00:10:27.440
And these peptides can be derived from four different target proteins that are transcriptionally independently regulated.
00:10:28.000 --> 00:10:58.799
Um, and it is that combinatorial optionality of expression in particular cells or lesions or patients that is beneficial to multi-targeting in the sense that ultimately it doesn't matter which combination or which individual target protein is expressed, as long as the peptide MHC complex is displayed on the cell surface, our binder can recognize it, and therefore our T cell engager can facilitate its mechanism of action.
00:10:59.039 --> 00:11:10.480
And that multi-targeting with a single binder or a single drug is actually the unique differentiation factor in the sort of molecular design behind um our approach attackbile.
00:11:11.360 --> 00:11:11.840
Yeah.
00:11:12.559 --> 00:11:13.840
Wow, thanks, Johanna.
00:11:14.080 --> 00:11:17.360
So I I'm sure that's not by coincidence, right?
00:11:17.519 --> 00:11:32.240
The uh the the fact that both targets combine in the same same way based on the Moety differences being inside or the epitopes being inside um and the outside.
00:11:32.720 --> 00:11:39.440
Um I'm curious to hear about the molecular design of your T cell engager as a multi-specific.
00:11:39.519 --> 00:11:41.679
So I I had uh Dr.
00:11:41.759 --> 00:11:54.559
JJU from AP Biosciences on recently, where we spoke about how design design decisions in uh multi-specifics can really impact their manufacturability down the line.
00:11:54.879 --> 00:12:06.720
And I'm curious to hear from you um in what ways you would agree with that and and perhaps what kind of decisions went into designing uh your lead access dBx01.
00:12:07.279 --> 00:12:18.879
Uh, did you think about ease of development and uh manufacturability at the onset and kind of how have those early decisions played out for you know compared to where you are right now?
00:12:19.120 --> 00:12:19.840
Yeah, yeah.
00:12:20.799 --> 00:12:27.120
Yeah, I think there are a number of ways to think about this question and how we ended up at the design of the T-cell engager that we currently have.
00:12:27.360 --> 00:12:37.919
If you think about this originally from the goal to try and address heterogeneity, you can do multi-targeting conceptually in three different ways, right?
00:12:38.159 --> 00:12:45.200
You could, in a simple way, say, I can just combine multiple drugs that recognize multiple targets.
00:12:45.360 --> 00:12:58.639
So you're making individual, in this case T cell engagers, traditional by specific molecules, and then you're combining a number of those together to address the multiple different targets that might be expressed on individual cells.
00:12:59.120 --> 00:13:04.799
Uh just last week we saw impressive examples of this approach by JJ.
00:13:05.039 --> 00:13:27.840
Um, the Monumental Six trial showed oppressive results by combining taclistomab and talchetomab in multiple myeloma, where two individual T cell engagers targeting BCMA and GPCR5D can combine and lead to very durable and oppressive responses, in part because you can cover a broad range of cells.
00:13:28.080 --> 00:13:35.759
Now, JJ in this case has the benefit that both of these T cell engagers are actually already approved drugs and other indications.
00:13:35.919 --> 00:13:44.720
So the idea of having to develop two drugs individually and then putting them together in a combination regimen for them is uh an option.
00:13:44.879 --> 00:13:52.000
For a novel biotech like us, this is essentially two programs and one, so not an attractive option to go after multi-targeting.
00:13:52.480 --> 00:13:59.840
The next way then is to say, okay, I'm going to combine this into multiple uh into a multi-specific molecule.
00:14:00.000 --> 00:14:03.600
I develop individual binders against multiple targets.
00:14:03.759 --> 00:14:13.840
I can develop those sequences for specificity and engineer them as much as I need to do that, and then throw them all together into a multi-specific molecule.
00:14:14.000 --> 00:14:17.919
That means cost of goods is now one molecule instead of two.
00:14:18.159 --> 00:14:30.159
At the same time, in particular for T cell engagers, you are starting to get really into questions about geometry and are both of these binders then equally active?
00:14:30.320 --> 00:14:35.200
What are the affinities, the affinity ratios, and compared to the CD3 arm?
00:14:35.600 --> 00:14:48.879
You need to have a very unique combination of properties of those individual binding moieties such that they can actually combine into a functional multi-specific molecule.
00:14:49.360 --> 00:14:59.919
In particular, when it comes to a T cell engager, we think about cross-linking two cell types together and creating an artificial immunological synapse.
00:15:00.080 --> 00:15:01.360
That sounds simple.
00:15:01.519 --> 00:15:15.840
The reality of the geometry of a molecule that needs to cross-link specific areas of two cells cannot be in the way of the signaling cascade or the effect of the T cell on the target cell at the end.
00:15:16.480 --> 00:15:24.000
The exact epitopes you're recognizing require different sterical arrangements that a molecule needs to be able to facilitate.
00:15:24.399 --> 00:15:39.919
So you come go into very deep linker optimization, epitope optimization that ends up with a molecule that can actually be quite complex and therefore ultimately hard to develop and manufacture.
00:15:40.320 --> 00:16:07.759
Therefore, we felt like this unique ability of finding combinations of target proteins that generate peptide MHD complexes that look identical, therefore, only requiring a single moiety for multi-targeting, is a very elegant way to make a simple bi-specific molecule that a priori is by default, therefore, potentially easier to manufacture.
00:16:08.559 --> 00:16:12.639
That was an elegant solution to the multi-targeting problem.
00:16:13.679 --> 00:16:16.879
Now I say in theory that makes it easy to manufacture.
00:16:17.039 --> 00:16:23.200
This is not where the story ends when the target class is peptide MHC complexes specifically.
00:16:23.360 --> 00:16:31.919
Because now, if you say, great, that's my target class, you have two options of how you can recognize the peptide MHC complex.
00:16:32.240 --> 00:16:51.600
You could do the maybe obvious traditional way, which would be just raise in monoclonal antibody against that and then derive your fab fragments, single chain FBs, whatever you need to do, and leverage decades' worth of antibody engineering strategies to make moieties that can be combined into a T cell engager.
00:16:52.879 --> 00:17:06.000
The other option is to take the natural interaction partner of a peptide MHC complex, which is a T cell receptor, another multimeric complex, and convert that into a single chain kind of moiety.
00:17:07.039 --> 00:17:28.319
Now, if we're thinking about this from a scientific specificity perspective, you would opt to go for T cell receptor because elegant literature has shown that there is a difference between the way a T cell receptor naturally clamps over the peptide MHD binding group as opposed to how an antibody would recognize this epitope.
00:17:28.640 --> 00:17:41.759
And it has been shown that generally speaking, T cell receptor recognition surfaces have a stronger contribution of the peptide than TCR mimetic antibodies.
00:17:41.839 --> 00:17:51.680
And that means the thing that drives specificity for that interaction is more prevalent to be recognized by TCRs.
00:17:52.319 --> 00:17:55.119
And therefore, ultimately you might have a cleaner molecule.
00:17:55.279 --> 00:18:04.319
Now, historically, T cell receptors, however, have had liabilities with respect to yields, thermal stability, aggregation propensity.
00:18:04.480 --> 00:18:07.039
So they're actually quite difficult to work with.
00:18:08.079 --> 00:18:22.559
And so those are theoretically the trade-offs that you have to make in choosing a design and an option as you start running a discovery campaign or lead optimization campaign to identify binders that are highly specific.
00:18:22.880 --> 00:18:41.119
And at Deck Bio, we've identified a way to stabilize soluble T cell receptors in a proprietary manner, such that we are bringing the developability profile back up into realms comparable to our monoclonal antibody controls that we run in the same processes.
00:18:41.759 --> 00:18:56.960
And that made it then with that solution in hand, which is unique to DEC Bio, quite obvious that the TCR route is the better and uh better way that allows us to generate highly specific finders.
00:18:57.599 --> 00:19:26.559
What are some of those trade-offs and what were the kind of development decisions that went into that to um you know, you mentioned those trade-offs like molecular stability with these T cell engagers and um walk me through some of those design choices if you if you could um yeah, I mean essentially the idea is that a T cell receptor is a multimer complex.
00:19:26.640 --> 00:19:29.839
You have to find ways to turn that into a single chain molecule.
00:19:30.000 --> 00:19:37.759
Similarly, like you have heavy and light chains that you're trying to turn into a single chain molecule in the form of a single gen F, for example.
00:19:38.000 --> 00:19:47.359
Um, and you need to make sure that that the different components that otherwise naturally are independent protein chains come together in an appropriate manner.
00:19:47.519 --> 00:19:57.839
Um, I can't go into the details of how exactly our proprietary stabilization technology ensures that that is the case in a more stable way than before.
00:19:58.000 --> 00:20:17.440
But what I can say is that our melting temperatures of the resulting binders are more akin to a monoclonal antibody like uh melting temperatures than they are to other T cell receptors, single-chain T cell receptors that do not contain the stabilization technology.
00:20:17.599 --> 00:20:54.960
Um, and that means that we can go from melting temperatures in the sort of 55, 56, 57 degree range uh to molecules that have melting temperatures above 60 degrees easily with um with very low aggregation propensity after forced um forced aggregation studies uh 37 degrees to weak stability um with less than 2% aggregation propensity, which uh is is quite different from what we have seen from non-stabilized TCR containing molecules.
00:20:55.440 --> 00:20:55.759
Yeah.
00:20:56.480 --> 00:20:56.880
Thank you.
00:20:57.039 --> 00:20:57.359
Thank you.
00:20:57.440 --> 00:20:58.319
That makes sense.
00:20:59.119 --> 00:21:05.599
Um and how how mature is this peptide AMHC targeting space that DeckBio is playing in?
00:21:05.759 --> 00:21:14.240
Like who has is there anybody even out there that you've been able to look to to kind of model uh this approach after, or are you breaking new ground?
00:21:14.400 --> 00:21:16.720
It kind of tell us a little bit about that.
00:21:17.119 --> 00:21:17.440
Yeah.
00:21:18.079 --> 00:21:22.720
Um so obviously T cell engagers as a modality are well established.
00:21:22.799 --> 00:21:25.200
They're 10 FDA-approved drugs at this point.
00:21:25.279 --> 00:21:36.240
So there is a lot of biology and mechanism of action, um, a lot of sequences and knowledge on the CD3 side, at least, uh, that we can learn from.
00:21:36.559 --> 00:21:48.240
The peptide MHC targeted T cell engager space is relatively niche, I would say, at this point, and still on the cutting edge of T cell engager development.
00:21:48.480 --> 00:21:55.119
However, we are not the only ones or the first ones to have thought about peptide MHCs as targeted.
00:21:55.359 --> 00:21:59.680
In fact, there is an approved agent, uh, Tibentifusp, marketed.
00:22:00.160 --> 00:22:29.440
Kim track um approved for U-val melanoma as actually the first solid tumor T cell engager of any kind, but they are recognizing a peptide MHC complex, a single one, uh derived from uh a protein called GP100, and that treats um uh uh eye cancer, U-val melanoma, um, a relatively niche indication with about a thousand patients in the US and EU eligible for this for this therapy.
00:22:29.839 --> 00:22:38.559
Tabentafusp is a non-half-life extended molecule, so that obviously comes with implication when it comes to dosing and PK.
00:22:38.720 --> 00:22:47.279
And that's certainly one of the pieces that we uh and others in the field have taken um, you know, a close look at.
00:22:47.440 --> 00:22:56.240
And I think the newer generations of molecules tend to be all half-life extended, and so is dbx01 and other pipeline assets for DeckBio.
00:22:57.039 --> 00:23:08.799
There are other companies in the peptide MHC targeting space uh with all their unique ways of how they're addressing a particular problem that they think in the field is important.
00:23:09.200 --> 00:23:15.680
Uh, CLASP, Crossbow, Inara, Evolve Immune are just uh a number of these companies.
00:23:16.000 --> 00:23:44.559
Um, however, the unique way of how we do multi-targeting and how that unlocks major solid tumor types for us, because 30 to 50 percent of patients with major solid tumor types, like non-small cell lung cancers, uh, head and neck cancers, and others along the aerodigestive tract express cell signature, that really creates a step change in the market opportunity for peptide MHT complexes.
00:23:44.799 --> 00:23:57.119
We're not talking about a thousand or 10,000 patients who might have a particular um neoantigen expressed or are a niche indication.
00:23:57.200 --> 00:24:07.680
We're targeting major solid tumor types with up to 120,000 biomarker positive patients for DBX01 in the Western world.
00:24:08.000 --> 00:24:20.240
So um, this is sort of how we're thinking to also address some of the market access barriers or market opportunity barriers uh that we often get asked about for peptide MHC targeting.
00:24:20.480 --> 00:24:26.160
Um, but at the end of the day, this is a tight-knit group of companies.
00:24:26.319 --> 00:24:29.839
Um, a rising tide lifts all ships in this case.
00:24:30.000 --> 00:24:35.359
It is overall niche enough that their success is ours, and vice versa.
00:24:35.519 --> 00:24:44.559
Um, but you know, we do learn from each other in in this field because we all uh want to achieve the same thing for patients after all.
00:24:45.200 --> 00:24:46.319
I love to hear that.
00:24:46.640 --> 00:24:47.680
Thanks, Johanna.
00:24:47.759 --> 00:24:48.640
I love to hear that.
00:24:49.359 --> 00:24:54.240
Um and I'm I wanted to ask a bit about your uh process development.
00:24:54.480 --> 00:24:58.240
And you know, I assume I know nothing about it, right?
00:24:58.319 --> 00:25:11.039
I do know some from conversations like this, but I wanted to ask you about your process development and what uh aspect of it has been the most challenging, whether that be upstream or downstream.
00:25:11.119 --> 00:25:27.680
Now, if I had to hazard a guess based on the conversations I've had, I'd guess that downstream usually presents more challenges in terms of purification and the way that multi-specifics like like this uh behave in in a process.
00:25:27.839 --> 00:25:39.680
But I I'd be curious to hear from you what maybe you were surprised by or you know what kind of was as expected, the kind of developmental challenges around this DBX01.
00:25:40.240 --> 00:25:40.799
Yeah.
00:25:41.119 --> 00:25:53.680
Um, I don't think I can share too many specifics about our process development challenges or solutions specifically, in the sense that uh formal CMC for lead asset is still upcoming.
00:25:53.759 --> 00:26:03.119
So we're as a next milestone actually approaching development candidate nomination at the end of this year, which will then trigger the formal CMC process.
00:26:03.599 --> 00:26:08.960
So um I hope that in a year from now I have a better answer for you to this question.
00:26:09.200 --> 00:26:14.559
At this point, I can speculate where we anticipate issues to occur.
00:26:14.720 --> 00:26:22.880
Um, and the first question for an asymmetric bi-specific molecule tends to be the yield, to be honest.
00:26:23.039 --> 00:26:24.960
So on the upstream side.
00:26:25.119 --> 00:26:35.519
Um, and again, from our research grade material productions, we're quite encouraged that yields will not be abysmal and actually an issue here.
00:26:35.759 --> 00:26:58.000
Um, but I cannot speak for data we have not generated, and we still need to go through the appropriate scaling process to confirm that in a process that it has, you know, GMP manufacturing um applications, we can replicate um the current finding that yield is not going to be an issue for us.
00:26:58.720 --> 00:27:01.200
Um yes.
00:27:01.440 --> 00:27:10.559
And I would assume, again, based on our current developability profile that we've generated based on research grade and research scale material.
00:27:10.720 --> 00:27:28.160
Um, I don't envision any super unique challenges that would come our way as compared to other bi-specific molecules, bi-specific T cell engagers, um, and on the basis of our stabilization technology that we call DBTB format.
00:27:28.400 --> 00:27:32.880
Um, but um the proof will be in the pudding when we actually do it.
00:27:33.359 --> 00:27:34.319
Right, right on.
00:27:34.400 --> 00:27:34.720
Thank you.
00:27:34.799 --> 00:27:40.240
I I understand the need to um you can't get into that too deep right now, but I appreciate you sharing what you did.
00:27:40.559 --> 00:27:58.559
Um can I ask what what you're doing at Deck Bio and specifically, you know, as the chief scientific officer, I know from other past conversations that um these CMC packages as you're approaching the clinic for the first time are far from a light lift.
00:27:58.640 --> 00:28:02.240
In fact, they're what keep folks like you up at night often.
00:28:02.319 --> 00:28:16.960
And I'm curious to hear if you can share a bit about how you and DeckBio are preparing for this CMC and uh um I and D kind of package, or if you're not at liberty to that's fine as well.
00:28:17.200 --> 00:28:44.000
Yeah, I think the the best way to answer that at this moment in time is to say that um we certainly need to build some internal capabilities with experienced um CMC experts, specifically a head of CMC, which is a role that will be part of our hiring um plan as we get closer to DC nomination, fundraise, and then execution of the CMC plans.
00:28:44.240 --> 00:28:58.559
And um, I look forward to partnering with that head of CMC as the subject matter expert in this area for them to really take that on and own the process and navigating the challenges that may or may not come our way.
00:28:59.759 --> 00:29:00.559
Got it.
00:29:00.799 --> 00:29:14.720
Yeah, by the by the time this comes out, it'll be right around I'm putting on an event about kind of preparing, you know, I guess maturing a process out of out of the lab and into uh GMP production.
00:29:15.200 --> 00:29:28.960
And is there any advice or um yeah, I guess advice you'd share about that as you as you kind of prepare to do that yourself at Deck Bio or some lessons you're learning along the way that you think our audience could appreciate?
00:29:29.359 --> 00:29:30.000
Yeah.
00:29:30.240 --> 00:29:52.480
Um maybe what I'll say here uh applies to sort of the relationship of biotech companies with CDMOs that would also apply in the preclinical space that I find spending most of my time in right now, where I work heavily with contract research organizations to operationalize our functional characterization of T cell engagers.
00:29:52.640 --> 00:30:13.839
And that is just um to be sort of careful on the biotech side, in particular if you're sort of a small fish in the pond, so to speak, as an early stage startup company with a small team where you wear a lot of hats and have to cover a lot of biologies and functions that you would otherwise have experts on in larger organizations.
00:30:14.640 --> 00:30:30.160
People make a lot of promises to you about what they can do, what they have done, what the timelines will be, and have also opinions about what regulators will be accepting as a package, as a release criterion, um or not.
00:30:30.400 --> 00:31:00.799
And I think my advice would be to try to the best ability you can to do your homework and actually, you know, want to see data that supports information that people give, examples of programs that they've run, examples of experiments they have done, potency assays that they have established, um, and and relying on precedence as well as doing your own homework of what's actually going to be required.
00:31:01.119 --> 00:31:12.799
And maybe also consider regulatory interaction a little bit earlier than later to be sure you're on the right track for release criteria, potency assay development, and other things.
00:31:13.039 --> 00:31:23.599
Again, for us, since we are not the very first peptide MHC targeted T cell engagers, there is a blueprint for us on what to follow and what might be required.
00:31:23.680 --> 00:31:29.279
Um and we certainly appreciate learning from others that are before us.
00:31:29.440 --> 00:31:32.319
Um, but that doesn't mean we don't need to do our own homework.
00:31:32.400 --> 00:31:39.359
And in particular, when it comes to rigorous sort of evaluation who the partner should be.
00:31:39.759 --> 00:31:54.000
Um, I would I would just say, you know, do your homework because for a startup company like Deck Bio, the CEMC contract is going to be the biggest contract we will have ever signed in the history of the company at that point in time, right?
00:31:54.240 --> 00:32:01.200
There might be bigger ones coming after or clinical, but at that moment in time, uh that will look like a big sum.
00:32:01.359 --> 00:32:05.519
So you want to be sure that, you know, in you are in this with a good partner.
00:32:06.000 --> 00:32:07.519
Yeah, absolutely.
00:32:07.680 --> 00:32:14.319
The importance of making the right kind of partnerships, uh, I hear it over and over time and time again.
00:32:14.480 --> 00:32:16.400
It's of critical importance.
00:32:16.559 --> 00:32:19.599
It could truly make or break a program, from what I've been told.
00:32:19.759 --> 00:32:23.599
And so um, best of luck in that in that decision-making process.
00:32:23.839 --> 00:32:30.400
Um I'm confident you guys will be happy and make good partnerships along the way.
00:32:30.720 --> 00:32:44.720
Um, and obviously you're charging ahead here, but thus far, I'd love to hear from you on uh what has been the biggest developmental challenge so far in in designing and developing DBX.
00:32:44.880 --> 00:32:48.480
Oh, one, uh especially as it relates to process development.
00:32:48.640 --> 00:33:00.960
Um and bonus points, if you also note what might have been the biggest challenge that you didn't anticipate, because I'm sure going into it you had an idea of what some pain points would be.
00:33:01.279 --> 00:33:02.960
Um yeah, yeah.
00:33:05.759 --> 00:33:09.039
Yeah, can take that in many different directions.
00:33:09.279 --> 00:33:37.839
Um, I would say a enormous challenge for us comes actually back to some of the earlier questions regarding to format and design of the molecule, not necessarily design in how many moieties do you want, or do you want half-life extension or not, um, but the intricacies of the exact format, of the affinities you need to achieve, the affinity ratios between the different arms.
00:33:38.559 --> 00:33:49.599
As much as there is precedence out there, I don't think the field understands exactly all the parameters that need to be tweaked in what the perfect design actually is.
00:33:49.839 --> 00:34:06.000
And a format that might work for one program or one target or one developer doesn't necessarily mean that it is if you just copy that and do this over in the next program, that it will actually give you the best format and design.
00:34:06.240 --> 00:34:11.199
And there's a lot of empirical testing that needs to get done.
00:34:11.440 --> 00:34:20.079
And developability is one aspect of these things that we need to test, but function and other pieces are actually contributing to this.
00:34:20.400 --> 00:34:33.039
But to do this empirical testing in the current environment where for biotech and startup companies like us, we still are dealing with you know financial restrictions.
00:34:33.280 --> 00:34:44.159
It is actually quite complex to figure out a smart development plan because you can't just brute force parallel track all sorts of testing.
00:34:44.320 --> 00:34:54.559
You have to, you know, pick a make decisions along the way and pick a few um educated guesses, I would say, in which direction to go.
00:34:54.800 --> 00:34:59.920
And navigating that from a scientific perspective is complex enough.
00:35:00.000 --> 00:35:22.960
But doing this in a financially restricted manner, um, where you can't do your full developability profile early on is the first step as you're designing a new format, um, purely out of financial reasons, um, is certainly a challenge that I didn't anticipate we would be in for such a long period of time.
00:35:23.039 --> 00:35:44.559
If I look at this at a at an ecosystem perspective, where in general, you know, um the uplifting of that developments on the MA and IPO side in later stage programs still take a while until um, you know, startups like us and biotechs are benefiting from the improvements in the overall financial climate.
00:35:46.239 --> 00:35:47.119
Got it.
00:35:47.519 --> 00:35:47.920
Got it.
00:35:48.000 --> 00:35:49.119
Thank you for sharing.
00:35:49.440 --> 00:35:59.199
Um and now taking a step back a little bit, I want to look more big picture at the state of the industry and some things um that you might be excited by.
00:35:59.360 --> 00:36:14.239
I'm curious to know, you know, looking out at the next couple of years, what developments are are uh exciting you, but also what do you see as the biggest challenges and opportunities in the space that you play in and in maybe in biotech more broadly?
00:36:14.559 --> 00:36:36.000
Yeah, I think um one of the things I'm rattling with a little bit as both an opportunity and a challenge, unfortunately, um is just the general concept of precision medicine, which is certainly where T cell engagers and specifically peptide MHC targeted T cell engagers uh fall into.
00:36:36.159 --> 00:36:40.000
Um that is on the one hand, because of the HLA restriction of our molecule.
00:36:40.079 --> 00:36:42.480
That's one biomarker that you need to consider.
00:36:42.960 --> 00:36:52.320
Um while for some that is already a hurdle, overall, I think this is the one that is that that is the one that we can navigate from a technical perspective.
00:36:52.400 --> 00:37:29.360
Um, but generally the idea of finding the right drug for the right patient to increase the likelihood that a patient will respond because they actually have the target or express whatever uh liability in the tumor you want to leverage, reducing the exposure of individuals who are just not expected to respond because they simply don't have the target or that particular mutation that you're going after, and just improvements in the time and cost to develop the drug if you're not drowning out your clinical signal by enriching for patients that are likely to respond, makes complete scientific sense to me.
00:37:29.440 --> 00:37:40.480
And I'm excited by the opportunities that we have to improve clinical care, even when it remains in an experimental status during clinical trials.
00:37:40.960 --> 00:37:53.280
At the same time, biomarker-selected drugs or clinical development either way, means that you're reducing your patient pool.
00:37:53.519 --> 00:38:26.320
And while I personally find it extremely valuable to have very good outcomes for, let's say, half the number of people rather than poor outcomes for a broader number of people, um, I still feel like there are a few gaps that we need to bridge when it comes to understanding for commercial teams and those who otherwise evaluate the value of a particular drug to really understand why making a difference, even if it is for a potentially smaller patient population, is important.
00:38:26.639 --> 00:38:33.440
So I think that's where a challenge and opportunity still remains in the precision medicine space.
00:38:33.599 --> 00:38:47.760
And more broadly, I guess what I'm excited about as a theme is to realize that more and more companies are starting to think about access of their drug.
00:38:48.320 --> 00:38:59.199
It's not just about making the most innovative, cool, fancy thing that in research hospitals like Dana Farber or MD Anderson can be administered and tested.
00:38:59.440 --> 00:39:09.039
But how do we make drugs that can be ultimately affordable with relatively low cost of goods, which is in part driven by good developability profiles?
00:39:09.119 --> 00:39:28.239
Um, but also learning more about um clinical safety profiles and how to manage adverse events that are often a class effect in community settings are topics that I hear being discussed more and more in companies that are working on solutions.
00:39:28.400 --> 00:39:37.760
I don't think today everything is equitable the way it should be, but it's definitely a topic that will improve clinical cancer care in the long run.
00:39:37.840 --> 00:39:44.639
And I'm excited to know that more and more people are thinking about these topics and then coming up with solutions for this.
00:39:45.199 --> 00:39:46.159
Yeah, me too.
00:39:46.480 --> 00:39:46.880
Me too.
00:39:46.960 --> 00:39:48.880
These are very promising developments.
00:39:49.039 --> 00:39:56.559
And like you said, I think focusing on that accessibility aspect, I mean, that I I hear that as well quite often.
00:39:56.719 --> 00:40:03.599
Um every conference I go to, I think that's that's the underlying theme of improving access, reducing the cost of these therapies.
00:40:03.679 --> 00:40:09.119
And I agree that the development profile, developability profile is a key part of that.
00:40:09.199 --> 00:40:16.880
And so kudos to you and your team for um doing your part and your in your scope of the of the industry to uh to do that.
00:40:17.119 --> 00:40:25.679
So, Johanna, now I want to ask you the question that I ask all my guests on this show, and that is big or small, how do you think we can better biopharma?
00:40:26.480 --> 00:40:28.719
Yeah, I love this question.
00:40:28.880 --> 00:40:31.119
Um, it lets me think big.
00:40:31.360 --> 00:40:44.239
So I get to talk about my favorite topic, which is I wish we could better biopharma by improving collaboration among different companies.
00:40:44.639 --> 00:41:03.760
Specifically, I'm thinking about the translational field where we actually have a lot of patient-derived data sets and characterization of tumor tissues that has been generated and is available.
00:41:04.159 --> 00:41:24.639
However, because we are not sharing these kinds of complex uh data sets, we are hindering our ability to identify biomarkers for particular uh drug, but also in general, in the broader sense, enabling uh new target discovery uh for innovative drugs of the future.
00:41:25.039 --> 00:41:31.280
We can probably spend a whole podcast episode talking about all the reasons why this is so difficult.
00:41:31.440 --> 00:41:40.159
There are obviously hurdles about not people wanting to not share confidential information or intellectual property, which I understand.
00:41:40.400 --> 00:41:47.920
There are technical hurdles about data annotation, sample annotation, even the methods with which data is acquired.
00:41:48.079 --> 00:41:49.360
I get all that.
00:41:49.679 --> 00:42:01.039
I still think it is challenges that are worth thinking about, how to overcome this when it comes to pooling data sets that exist of many, many patients.
00:42:01.280 --> 00:42:04.800
Let's just call it the standard of care comparator arms.
00:42:04.880 --> 00:42:28.960
If we could just have that as a tool available to the ecosystem for better drug development in the future, that would be an immense step forward that also honors the idea that patients selflessly actually allow us to collect biopsies in other specimens and make them available for research purposes.
00:42:29.199 --> 00:42:35.760
So I think the ecosystem would be better for more collaboration on this front.
00:42:35.920 --> 00:42:41.679
Um spanning from discovery to clinical development and care delivery.
00:42:42.960 --> 00:42:46.239
Yeah, I completely, I completely agree with that, Johanna.
00:42:46.559 --> 00:42:47.039
Absolutely.
00:42:47.119 --> 00:42:49.199
And I hope we get to see more of that.
00:42:49.440 --> 00:43:02.320
Um, you know, there's promising developments on the horizon, but I do hope as we move forward, especially in the age of AI, and that data sharing and insight sharing uh continues to grow alongside that.
00:43:02.719 --> 00:43:05.440
Um yeah, it's a great answer to that question.
00:43:05.599 --> 00:43:10.880
And thank you so much for joining me on this episode of Better Biopharma, the official podcast of Bioprocess Online.
00:43:11.039 --> 00:43:13.039
And thank you out there in the audience for tuning in.
00:43:13.199 --> 00:43:14.079
We'll see you next time.
00:00:17.120 --> 00:00:21.440
Hello and welcome back to Better BioForma, the official podcast of Bioprocess Online.
00:00:21.760 --> 00:00:32.719
I'm your host, Tyler Menechello, and on this episode, I'm joined by Johanna Kaufman, Chief Scientific Officer at Deck Bio, a company developing multipeptide MHC targeted T cell engagers for solid tumors.
00:00:32.880 --> 00:00:34.960
Johanna, thank you so much for joining me today.
00:00:35.520 --> 00:00:36.399
Thank you for having me.
00:00:36.479 --> 00:00:39.520
Thank you for the invitation, Tyler, and I look forward to the conversation.
00:00:39.920 --> 00:00:40.640
Likewise.
00:00:40.799 --> 00:00:52.880
So to start, why don't you tell our audience a bit about Deck Bio and what the heck I mean when I say multipeptide MHC-targeted T cell engagers, which I don't think I could say five times fast, even if I if I I know it's a mouthful.
00:00:54.079 --> 00:00:54.399
Yes.
00:00:54.560 --> 00:01:00.079
So DeckBio is a pre-seed stage startup company based in Cambridge, Massachusetts.
00:01:00.320 --> 00:01:05.599
And we are, like you said, developing novel T cell engagers for solid tumors.
00:01:05.840 --> 00:01:22.959
T cell engagers are an up-and-coming modality in the immunooncology field, and they're essentially cross-linking and affector immune cell T cells specifically to cancer cells, and then mediating essentially cell killing.
00:01:23.680 --> 00:01:39.519
This modality is well established in hematological malignancies or blood cancers, but translation into the solid tumor space, which is actually 95% of all cancers, has been lagging behind and is so far restricted to very niche indications.
00:01:39.680 --> 00:02:00.799
And it is our mission at Dech Bio to take the learnings from the clinical and design perspectives of T cell engagers together with the biology in solid tumors and address this unmet need and therefore bringing T cell engagers to broader patient populations with solid tumors specifically.
00:02:01.120 --> 00:02:08.479
We're doing this in the way of choosing peptide MHCs as a target class.
00:02:08.719 --> 00:02:18.000
The reason behind this is that traditional cell surface targets often have residual expression on healthy cells.
00:02:18.159 --> 00:02:32.159
And while that may be okay for other modalities like antibody drug conjugates, for example, as long as you have the right dosing window, for very potent modalities like T cell engagers, that is actually a safety liability.
00:02:32.400 --> 00:02:41.439
So we want to choose targets that mark the cancer cells specifically that are expressed as exclusively in tumors as possible.
00:02:41.680 --> 00:02:52.319
Now, the problem is that most of those targets are actually intracellularly expressed and are not naturally easily accessible to biologics like T cell engagers.
00:02:52.479 --> 00:02:59.599
And to overcome this hurdle, we are targeting these molecules through peptide MHC complexes.
00:02:59.840 --> 00:03:14.879
These are generated as part of natural immune surveillance mechanisms of the body, where essentially every intracellular protein gets chopped up into peptides and sampled on the cell surface and displayed in the form of these peptide MHC complexes.
00:03:14.960 --> 00:03:18.879
And that's what we're using as a target class, as the foundation.
00:03:19.039 --> 00:03:22.800
And then we're taking this approach to the next level through multi-targeting.
00:03:22.960 --> 00:03:30.479
And I'm sure we'll get into the rationale and the ways we do this in a little bit more detail as part of this conversation.
00:03:30.719 --> 00:03:45.120
We are currently focusing our efforts on our lead asset, DBX01, which recognizes a signature displayed in the most prevalent HLA allele of the Western world, HLA A0201.
00:03:45.520 --> 00:04:00.800
And we are driving the RD process towards entering the clinic in 2028 with additional pipeline programs that leverage the same multi-peptidemia C-targeted concept in the context of additional HLA alleles.
00:04:02.000 --> 00:04:02.800
Great, Johanna.
00:04:02.960 --> 00:04:07.439
Thank you so much for that succinct and uh comprehensive overview there.
00:04:07.599 --> 00:04:08.560
I appreciate that.
00:04:08.719 --> 00:04:18.000
And you're right, we will be getting into uh the rationale behind multi-targeting and kind of the structural um molecular design underpinning that.
00:04:18.160 --> 00:04:20.240
So I'd love to hear from you, you know.
00:04:20.319 --> 00:04:30.560
Yeah, of course you hear about the challenges of multi-targeting, but I before we get into that, I want to hear from you about the rationale to choose multi-targeting.
00:04:30.639 --> 00:04:33.759
Why do you think that that's necessary from a therapeutic standpoint?
00:04:33.839 --> 00:04:43.920
And is it something you mentioned that uh this we have T cell engagers in the blood cancer space, not not really much in the solid tumor space.
00:04:44.560 --> 00:04:56.319
I've heard from others that because of the the makeup of antigens that's associated with solid tumors, that uh it kind of lends itself to to multi-targeting or more or less demands it.
00:04:56.399 --> 00:05:01.279
So yeah, I'd be curious to hear um why you think it's necessary from a therapeutic standpoint.
00:05:01.839 --> 00:05:09.519
Yeah, I think the main reason why multi-targeting is necessary is heterogeneity.
00:05:10.000 --> 00:05:29.519
You know, and when we draw up mechanism of action for a drug, we generally depict the tumor as a homogeneous, uniform mass where every cell within that tumor, every patient seemingly expresses a particular target or has a particular aberration that a drug can address.
00:05:29.680 --> 00:05:37.839
And then therefore, theoretically, every one of these cancer cells or every patient in a cohort will respond to therapy and be cured.
00:05:38.079 --> 00:05:41.600
The reality is that that's not how this really works.
00:05:41.759 --> 00:05:46.560
We actually have quite uh heterogeneous tumors with different parameters.
00:05:46.720 --> 00:05:54.319
And you can think about this at the level of different individuals, different patients within the same tumor type.
00:05:54.639 --> 00:06:04.639
So if we both were in the unfortunate situation to have, let's say, non-small cell lung cancer, your molecular makeup might not be identical to mine.
00:06:04.800 --> 00:06:13.439
It might also not be that every lesion within a patient's body has the same makeup, or every cell, even within a tumor lesion.
00:06:13.600 --> 00:06:19.759
And there are different types of targets that might be expressed, or at least at different levels.
00:06:20.160 --> 00:06:27.600
This phenomenon of heterogeneity is much more pronounced in solid tumors than it is in blood cancers.
00:06:27.759 --> 00:06:36.639
And so you're right, this problem is much more pronounced and a lot much larger hurdle that needs to be addressed in solid tumors.
00:06:36.959 --> 00:06:55.680
So if we're focusing on one aspect of this heterogeneity, which is target expression, we often realize that target is expressed at different levels in different cells or patients, or only a fraction of the cells actually expresses or displays a target.
00:06:55.839 --> 00:07:11.680
So if I make a T cell engager against a single target, the traditional bispecific T cell engager, I might only eliminate a subset of the cells, the remainder grows out, and I ultimately have not achieved a desirable outcome for that patient.
00:07:12.000 --> 00:07:24.959
By going after multiple targets at the same time, I have a higher chance that I can cover the different types of subsets of cells within a tumor nodule or different lesions or different patients.
00:07:25.120 --> 00:07:35.279
And therefore, I can increase the likelihood of response, I can increase the number of patients that might respond to a therapy.
00:07:35.519 --> 00:07:40.879
Um, and that is obviously beneficial from an outcome, clinical outcome as well as commercial perspective.
00:07:41.040 --> 00:07:52.959
And that is why we strongly believe that multi-targeting in general, whether you do this through peptide MHC complexes or otherwise, and whether you do it in our unique way or otherwise.
00:07:53.360 --> 00:08:01.120
But the concept of multi-targeting, we believe, is extremely important for durable responses and broad patient populations.
00:08:01.759 --> 00:08:02.160
Yeah.
00:08:02.639 --> 00:08:04.000
Thank you, Johanna.
00:08:04.480 --> 00:08:13.759
And I was particularly fascinated by um your ability to target a single binding domain with your T cell engagers.
00:08:13.839 --> 00:08:21.680
And I'm curious if you could explain how you are able to do that to uh target multiple peptide MHC complexes with a single binding domain.
00:08:21.759 --> 00:08:25.839
I don't think that's common in the T cell engager kind of landscape.
00:08:26.079 --> 00:08:28.480
Correct me if I'm wrong, but that is quite new.
00:08:29.040 --> 00:08:32.080
That's accurate, at least for multi-targeted uh T cell engagers.
00:08:32.240 --> 00:08:37.679
So our molecule is uh in some ways a very simple bi-specific molecule.
00:08:37.840 --> 00:08:42.000
One arm obviously targets the T cell side of things, CD3 in our case.
00:08:42.159 --> 00:08:44.000
We're not reinventing the wheel here.
00:08:44.159 --> 00:08:54.399
The unique aspect is the cancer-engaging arm, which is a single binding moiety that recognizes multiple targets.
00:08:54.720 --> 00:09:29.360
The reason why this is possible is that during target identification, we are identifying signatures of multiple target proteins that are expressed specifically in cancer, not in healthy cells, that derive peptides that are either identical to each other or highly similar to each other, such that these peptides get loaded into the same HLA allele and can therefore then be recognized by a single binder.
00:09:29.759 --> 00:09:46.720
The key here is that these different peptides, as they lay in the peptide MHC binding groove, display the same or almost the same surface topology or 3D epitope, if you will, that a binder can recognize.
00:09:47.039 --> 00:10:02.720
Any sequence differences of these peptides are pointing inwards into the peptide MHD binding groove inside into the complex, such that they do not contribute to that surface topology.
00:10:02.799 --> 00:10:18.799
So, from the perspective of the binder, it actually looks like the same thing, except for that it is actually in the case of our lead acid DBX01, two peptides that get loaded into the same HLA allele, again in our case, AO201.
00:10:18.960 --> 00:10:27.440
And these peptides can be derived from four different target proteins that are transcriptionally independently regulated.
00:10:28.000 --> 00:10:58.799
Um, and it is that combinatorial optionality of expression in particular cells or lesions or patients that is beneficial to multi-targeting in the sense that ultimately it doesn't matter which combination or which individual target protein is expressed, as long as the peptide MHC complex is displayed on the cell surface, our binder can recognize it, and therefore our T cell engager can facilitate its mechanism of action.
00:10:59.039 --> 00:11:10.480
And that multi-targeting with a single binder or a single drug is actually the unique differentiation factor in the sort of molecular design behind um our approach attackbile.
00:11:11.360 --> 00:11:11.840
Yeah.
00:11:12.559 --> 00:11:13.840
Wow, thanks, Johanna.
00:11:14.080 --> 00:11:17.360
So I I'm sure that's not by coincidence, right?
00:11:17.519 --> 00:11:32.240
The uh the the fact that both targets combine in the same same way based on the Moety differences being inside or the epitopes being inside um and the outside.
00:11:32.720 --> 00:11:39.440
Um I'm curious to hear about the molecular design of your T cell engager as a multi-specific.
00:11:39.519 --> 00:11:41.679
So I I had uh Dr.
00:11:41.759 --> 00:11:54.559
JJU from AP Biosciences on recently, where we spoke about how design design decisions in uh multi-specifics can really impact their manufacturability down the line.
00:11:54.879 --> 00:12:06.720
And I'm curious to hear from you um in what ways you would agree with that and and perhaps what kind of decisions went into designing uh your lead access dBx01.
00:12:07.279 --> 00:12:18.879
Uh, did you think about ease of development and uh manufacturability at the onset and kind of how have those early decisions played out for you know compared to where you are right now?
00:12:19.120 --> 00:12:19.840
Yeah, yeah.
00:12:20.799 --> 00:12:27.120
Yeah, I think there are a number of ways to think about this question and how we ended up at the design of the T-cell engager that we currently have.
00:12:27.360 --> 00:12:37.919
If you think about this originally from the goal to try and address heterogeneity, you can do multi-targeting conceptually in three different ways, right?
00:12:38.159 --> 00:12:45.200
You could, in a simple way, say, I can just combine multiple drugs that recognize multiple targets.
00:12:45.360 --> 00:12:58.639
So you're making individual, in this case T cell engagers, traditional by specific molecules, and then you're combining a number of those together to address the multiple different targets that might be expressed on individual cells.
00:12:59.120 --> 00:13:04.799
Uh just last week we saw impressive examples of this approach by JJ.
00:13:05.039 --> 00:13:27.840
Um, the Monumental Six trial showed oppressive results by combining taclistomab and talchetomab in multiple myeloma, where two individual T cell engagers targeting BCMA and GPCR5D can combine and lead to very durable and oppressive responses, in part because you can cover a broad range of cells.
00:13:28.080 --> 00:13:35.759
Now, JJ in this case has the benefit that both of these T cell engagers are actually already approved drugs and other indications.
00:13:35.919 --> 00:13:44.720
So the idea of having to develop two drugs individually and then putting them together in a combination regimen for them is uh an option.
00:13:44.879 --> 00:13:52.000
For a novel biotech like us, this is essentially two programs and one, so not an attractive option to go after multi-targeting.
00:13:52.480 --> 00:13:59.840
The next way then is to say, okay, I'm going to combine this into multiple uh into a multi-specific molecule.
00:14:00.000 --> 00:14:03.600
I develop individual binders against multiple targets.
00:14:03.759 --> 00:14:13.840
I can develop those sequences for specificity and engineer them as much as I need to do that, and then throw them all together into a multi-specific molecule.
00:14:14.000 --> 00:14:17.919
That means cost of goods is now one molecule instead of two.
00:14:18.159 --> 00:14:30.159
At the same time, in particular for T cell engagers, you are starting to get really into questions about geometry and are both of these binders then equally active?
00:14:30.320 --> 00:14:35.200
What are the affinities, the affinity ratios, and compared to the CD3 arm?
00:14:35.600 --> 00:14:48.879
You need to have a very unique combination of properties of those individual binding moieties such that they can actually combine into a functional multi-specific molecule.
00:14:49.360 --> 00:14:59.919
In particular, when it comes to a T cell engager, we think about cross-linking two cell types together and creating an artificial immunological synapse.
00:15:00.080 --> 00:15:01.360
That sounds simple.
00:15:01.519 --> 00:15:15.840
The reality of the geometry of a molecule that needs to cross-link specific areas of two cells cannot be in the way of the signaling cascade or the effect of the T cell on the target cell at the end.
00:15:16.480 --> 00:15:24.000
The exact epitopes you're recognizing require different sterical arrangements that a molecule needs to be able to facilitate.
00:15:24.399 --> 00:15:39.919
So you come go into very deep linker optimization, epitope optimization that ends up with a molecule that can actually be quite complex and therefore ultimately hard to develop and manufacture.
00:15:40.320 --> 00:16:07.759
Therefore, we felt like this unique ability of finding combinations of target proteins that generate peptide MHD complexes that look identical, therefore, only requiring a single moiety for multi-targeting, is a very elegant way to make a simple bi-specific molecule that a priori is by default, therefore, potentially easier to manufacture.
00:16:08.559 --> 00:16:12.639
That was an elegant solution to the multi-targeting problem.
00:16:13.679 --> 00:16:16.879
Now I say in theory that makes it easy to manufacture.
00:16:17.039 --> 00:16:23.200
This is not where the story ends when the target class is peptide MHC complexes specifically.
00:16:23.360 --> 00:16:31.919
Because now, if you say, great, that's my target class, you have two options of how you can recognize the peptide MHC complex.
00:16:32.240 --> 00:16:51.600
You could do the maybe obvious traditional way, which would be just raise in monoclonal antibody against that and then derive your fab fragments, single chain FBs, whatever you need to do, and leverage decades' worth of antibody engineering strategies to make moieties that can be combined into a T cell engager.
00:16:52.879 --> 00:17:06.000
The other option is to take the natural interaction partner of a peptide MHC complex, which is a T cell receptor, another multimeric complex, and convert that into a single chain kind of moiety.
00:17:07.039 --> 00:17:28.319
Now, if we're thinking about this from a scientific specificity perspective, you would opt to go for T cell receptor because elegant literature has shown that there is a difference between the way a T cell receptor naturally clamps over the peptide MHD binding group as opposed to how an antibody would recognize this epitope.
00:17:28.640 --> 00:17:41.759
And it has been shown that generally speaking, T cell receptor recognition surfaces have a stronger contribution of the peptide than TCR mimetic antibodies.
00:17:41.839 --> 00:17:51.680
And that means the thing that drives specificity for that interaction is more prevalent to be recognized by TCRs.
00:17:52.319 --> 00:17:55.119
And therefore, ultimately you might have a cleaner molecule.
00:17:55.279 --> 00:18:04.319
Now, historically, T cell receptors, however, have had liabilities with respect to yields, thermal stability, aggregation propensity.
00:18:04.480 --> 00:18:07.039
So they're actually quite difficult to work with.
00:18:08.079 --> 00:18:22.559
And so those are theoretically the trade-offs that you have to make in choosing a design and an option as you start running a discovery campaign or lead optimization campaign to identify binders that are highly specific.
00:18:22.880 --> 00:18:41.119
And at Deck Bio, we've identified a way to stabilize soluble T cell receptors in a proprietary manner, such that we are bringing the developability profile back up into realms comparable to our monoclonal antibody controls that we run in the same processes.
00:18:41.759 --> 00:18:56.960
And that made it then with that solution in hand, which is unique to DEC Bio, quite obvious that the TCR route is the better and uh better way that allows us to generate highly specific finders.
00:18:57.599 --> 00:19:26.559
What are some of those trade-offs and what were the kind of development decisions that went into that to um you know, you mentioned those trade-offs like molecular stability with these T cell engagers and um walk me through some of those design choices if you if you could um yeah, I mean essentially the idea is that a T cell receptor is a multimer complex.
00:19:26.640 --> 00:19:29.839
You have to find ways to turn that into a single chain molecule.
00:19:30.000 --> 00:19:37.759
Similarly, like you have heavy and light chains that you're trying to turn into a single chain molecule in the form of a single gen F, for example.
00:19:38.000 --> 00:19:47.359
Um, and you need to make sure that that the different components that otherwise naturally are independent protein chains come together in an appropriate manner.
00:19:47.519 --> 00:19:57.839
Um, I can't go into the details of how exactly our proprietary stabilization technology ensures that that is the case in a more stable way than before.
00:19:58.000 --> 00:20:17.440
But what I can say is that our melting temperatures of the resulting binders are more akin to a monoclonal antibody like uh melting temperatures than they are to other T cell receptors, single-chain T cell receptors that do not contain the stabilization technology.
00:20:17.599 --> 00:20:54.960
Um, and that means that we can go from melting temperatures in the sort of 55, 56, 57 degree range uh to molecules that have melting temperatures above 60 degrees easily with um with very low aggregation propensity after forced um forced aggregation studies uh 37 degrees to weak stability um with less than 2% aggregation propensity, which uh is is quite different from what we have seen from non-stabilized TCR containing molecules.
00:20:55.440 --> 00:20:55.759
Yeah.
00:20:56.480 --> 00:20:56.880
Thank you.
00:20:57.039 --> 00:20:57.359
Thank you.
00:20:57.440 --> 00:20:58.319
That makes sense.
00:20:59.119 --> 00:21:05.599
Um and how how mature is this peptide AMHC targeting space that DeckBio is playing in?
00:21:05.759 --> 00:21:14.240
Like who has is there anybody even out there that you've been able to look to to kind of model uh this approach after, or are you breaking new ground?
00:21:14.400 --> 00:21:16.720
It kind of tell us a little bit about that.
00:21:17.119 --> 00:21:17.440
Yeah.
00:21:18.079 --> 00:21:22.720
Um so obviously T cell engagers as a modality are well established.
00:21:22.799 --> 00:21:25.200
They're 10 FDA-approved drugs at this point.
00:21:25.279 --> 00:21:36.240
So there is a lot of biology and mechanism of action, um, a lot of sequences and knowledge on the CD3 side, at least, uh, that we can learn from.
00:21:36.559 --> 00:21:48.240
The peptide MHC targeted T cell engager space is relatively niche, I would say, at this point, and still on the cutting edge of T cell engager development.
00:21:48.480 --> 00:21:55.119
However, we are not the only ones or the first ones to have thought about peptide MHCs as targeted.
00:21:55.359 --> 00:21:59.680
In fact, there is an approved agent, uh, Tibentifusp, marketed.
00:22:00.160 --> 00:22:29.440
Kim track um approved for U-val melanoma as actually the first solid tumor T cell engager of any kind, but they are recognizing a peptide MHC complex, a single one, uh derived from uh a protein called GP100, and that treats um uh uh eye cancer, U-val melanoma, um, a relatively niche indication with about a thousand patients in the US and EU eligible for this for this therapy.
00:22:29.839 --> 00:22:38.559
Tabentafusp is a non-half-life extended molecule, so that obviously comes with implication when it comes to dosing and PK.
00:22:38.720 --> 00:22:47.279
And that's certainly one of the pieces that we uh and others in the field have taken um, you know, a close look at.
00:22:47.440 --> 00:22:56.240
And I think the newer generations of molecules tend to be all half-life extended, and so is dbx01 and other pipeline assets for DeckBio.
00:22:57.039 --> 00:23:08.799
There are other companies in the peptide MHC targeting space uh with all their unique ways of how they're addressing a particular problem that they think in the field is important.
00:23:09.200 --> 00:23:15.680
Uh, CLASP, Crossbow, Inara, Evolve Immune are just uh a number of these companies.
00:23:16.000 --> 00:23:44.559
Um, however, the unique way of how we do multi-targeting and how that unlocks major solid tumor types for us, because 30 to 50 percent of patients with major solid tumor types, like non-small cell lung cancers, uh, head and neck cancers, and others along the aerodigestive tract express cell signature, that really creates a step change in the market opportunity for peptide MHT complexes.
00:23:44.799 --> 00:23:57.119
We're not talking about a thousand or 10,000 patients who might have a particular um neoantigen expressed or are a niche indication.
00:23:57.200 --> 00:24:07.680
We're targeting major solid tumor types with up to 120,000 biomarker positive patients for DBX01 in the Western world.
00:24:08.000 --> 00:24:20.240
So um, this is sort of how we're thinking to also address some of the market access barriers or market opportunity barriers uh that we often get asked about for peptide MHC targeting.
00:24:20.480 --> 00:24:26.160
Um, but at the end of the day, this is a tight-knit group of companies.
00:24:26.319 --> 00:24:29.839
Um, a rising tide lifts all ships in this case.
00:24:30.000 --> 00:24:35.359
It is overall niche enough that their success is ours, and vice versa.
00:24:35.519 --> 00:24:44.559
Um, but you know, we do learn from each other in in this field because we all uh want to achieve the same thing for patients after all.
00:24:45.200 --> 00:24:46.319
I love to hear that.
00:24:46.640 --> 00:24:47.680
Thanks, Johanna.
00:24:47.759 --> 00:24:48.640
I love to hear that.
00:24:49.359 --> 00:24:54.240
Um and I'm I wanted to ask a bit about your uh process development.
00:24:54.480 --> 00:24:58.240
And you know, I assume I know nothing about it, right?
00:24:58.319 --> 00:25:11.039
I do know some from conversations like this, but I wanted to ask you about your process development and what uh aspect of it has been the most challenging, whether that be upstream or downstream.
00:25:11.119 --> 00:25:27.680
Now, if I had to hazard a guess based on the conversations I've had, I'd guess that downstream usually presents more challenges in terms of purification and the way that multi-specifics like like this uh behave in in a process.
00:25:27.839 --> 00:25:39.680
But I I'd be curious to hear from you what maybe you were surprised by or you know what kind of was as expected, the kind of developmental challenges around this DBX01.
00:25:40.240 --> 00:25:40.799
Yeah.
00:25:41.119 --> 00:25:53.680
Um, I don't think I can share too many specifics about our process development challenges or solutions specifically, in the sense that uh formal CMC for lead asset is still upcoming.
00:25:53.759 --> 00:26:03.119
So we're as a next milestone actually approaching development candidate nomination at the end of this year, which will then trigger the formal CMC process.
00:26:03.599 --> 00:26:08.960
So um I hope that in a year from now I have a better answer for you to this question.
00:26:09.200 --> 00:26:14.559
At this point, I can speculate where we anticipate issues to occur.
00:26:14.720 --> 00:26:22.880
Um, and the first question for an asymmetric bi-specific molecule tends to be the yield, to be honest.
00:26:23.039 --> 00:26:24.960
So on the upstream side.
00:26:25.119 --> 00:26:35.519
Um, and again, from our research grade material productions, we're quite encouraged that yields will not be abysmal and actually an issue here.
00:26:35.759 --> 00:26:58.000
Um, but I cannot speak for data we have not generated, and we still need to go through the appropriate scaling process to confirm that in a process that it has, you know, GMP manufacturing um applications, we can replicate um the current finding that yield is not going to be an issue for us.
00:26:58.720 --> 00:27:01.200
Um yes.
00:27:01.440 --> 00:27:10.559
And I would assume, again, based on our current developability profile that we've generated based on research grade and research scale material.
00:27:10.720 --> 00:27:28.160
Um, I don't envision any super unique challenges that would come our way as compared to other bi-specific molecules, bi-specific T cell engagers, um, and on the basis of our stabilization technology that we call DBTB format.
00:27:28.400 --> 00:27:32.880
Um, but um the proof will be in the pudding when we actually do it.
00:27:33.359 --> 00:27:34.319
Right, right on.
00:27:34.400 --> 00:27:34.720
Thank you.
00:27:34.799 --> 00:27:40.240
I I understand the need to um you can't get into that too deep right now, but I appreciate you sharing what you did.
00:27:40.559 --> 00:27:58.559
Um can I ask what what you're doing at Deck Bio and specifically, you know, as the chief scientific officer, I know from other past conversations that um these CMC packages as you're approaching the clinic for the first time are far from a light lift.
00:27:58.640 --> 00:28:02.240
In fact, they're what keep folks like you up at night often.
00:28:02.319 --> 00:28:16.960
And I'm curious to hear if you can share a bit about how you and DeckBio are preparing for this CMC and uh um I and D kind of package, or if you're not at liberty to that's fine as well.
00:28:17.200 --> 00:28:44.000
Yeah, I think the the best way to answer that at this moment in time is to say that um we certainly need to build some internal capabilities with experienced um CMC experts, specifically a head of CMC, which is a role that will be part of our hiring um plan as we get closer to DC nomination, fundraise, and then execution of the CMC plans.
00:28:44.240 --> 00:28:58.559
And um, I look forward to partnering with that head of CMC as the subject matter expert in this area for them to really take that on and own the process and navigating the challenges that may or may not come our way.
00:28:59.759 --> 00:29:00.559
Got it.
00:29:00.799 --> 00:29:14.720
Yeah, by the by the time this comes out, it'll be right around I'm putting on an event about kind of preparing, you know, I guess maturing a process out of out of the lab and into uh GMP production.
00:29:15.200 --> 00:29:28.960
And is there any advice or um yeah, I guess advice you'd share about that as you as you kind of prepare to do that yourself at Deck Bio or some lessons you're learning along the way that you think our audience could appreciate?
00:29:29.359 --> 00:29:30.000
Yeah.
00:29:30.240 --> 00:29:52.480
Um maybe what I'll say here uh applies to sort of the relationship of biotech companies with CDMOs that would also apply in the preclinical space that I find spending most of my time in right now, where I work heavily with contract research organizations to operationalize our functional characterization of T cell engagers.
00:29:52.640 --> 00:30:13.839
And that is just um to be sort of careful on the biotech side, in particular if you're sort of a small fish in the pond, so to speak, as an early stage startup company with a small team where you wear a lot of hats and have to cover a lot of biologies and functions that you would otherwise have experts on in larger organizations.
00:30:14.640 --> 00:30:30.160
People make a lot of promises to you about what they can do, what they have done, what the timelines will be, and have also opinions about what regulators will be accepting as a package, as a release criterion, um or not.
00:30:30.400 --> 00:31:00.799
And I think my advice would be to try to the best ability you can to do your homework and actually, you know, want to see data that supports information that people give, examples of programs that they've run, examples of experiments they have done, potency assays that they have established, um, and and relying on precedence as well as doing your own homework of what's actually going to be required.
00:31:01.119 --> 00:31:12.799
And maybe also consider regulatory interaction a little bit earlier than later to be sure you're on the right track for release criteria, potency assay development, and other things.
00:31:13.039 --> 00:31:23.599
Again, for us, since we are not the very first peptide MHC targeted T cell engagers, there is a blueprint for us on what to follow and what might be required.
00:31:23.680 --> 00:31:29.279
Um and we certainly appreciate learning from others that are before us.
00:31:29.440 --> 00:31:32.319
Um, but that doesn't mean we don't need to do our own homework.
00:31:32.400 --> 00:31:39.359
And in particular, when it comes to rigorous sort of evaluation who the partner should be.
00:31:39.759 --> 00:31:54.000
Um, I would I would just say, you know, do your homework because for a startup company like Deck Bio, the CEMC contract is going to be the biggest contract we will have ever signed in the history of the company at that point in time, right?
00:31:54.240 --> 00:32:01.200
There might be bigger ones coming after or clinical, but at that moment in time, uh that will look like a big sum.
00:32:01.359 --> 00:32:05.519
So you want to be sure that, you know, in you are in this with a good partner.
00:32:06.000 --> 00:32:07.519
Yeah, absolutely.
00:32:07.680 --> 00:32:14.319
The importance of making the right kind of partnerships, uh, I hear it over and over time and time again.
00:32:14.480 --> 00:32:16.400
It's of critical importance.
00:32:16.559 --> 00:32:19.599
It could truly make or break a program, from what I've been told.
00:32:19.759 --> 00:32:23.599
And so um, best of luck in that in that decision-making process.
00:32:23.839 --> 00:32:30.400
Um I'm confident you guys will be happy and make good partnerships along the way.
00:32:30.720 --> 00:32:44.720
Um, and obviously you're charging ahead here, but thus far, I'd love to hear from you on uh what has been the biggest developmental challenge so far in in designing and developing DBX.
00:32:44.880 --> 00:32:48.480
Oh, one, uh especially as it relates to process development.
00:32:48.640 --> 00:33:00.960
Um and bonus points, if you also note what might have been the biggest challenge that you didn't anticipate, because I'm sure going into it you had an idea of what some pain points would be.
00:33:01.279 --> 00:33:02.960
Um yeah, yeah.
00:33:05.759 --> 00:33:09.039
Yeah, can take that in many different directions.
00:33:09.279 --> 00:33:37.839
Um, I would say a enormous challenge for us comes actually back to some of the earlier questions regarding to format and design of the molecule, not necessarily design in how many moieties do you want, or do you want half-life extension or not, um, but the intricacies of the exact format, of the affinities you need to achieve, the affinity ratios between the different arms.
00:33:38.559 --> 00:33:49.599
As much as there is precedence out there, I don't think the field understands exactly all the parameters that need to be tweaked in what the perfect design actually is.
00:33:49.839 --> 00:34:06.000
And a format that might work for one program or one target or one developer doesn't necessarily mean that it is if you just copy that and do this over in the next program, that it will actually give you the best format and design.
00:34:06.240 --> 00:34:11.199
And there's a lot of empirical testing that needs to get done.
00:34:11.440 --> 00:34:20.079
And developability is one aspect of these things that we need to test, but function and other pieces are actually contributing to this.
00:34:20.400 --> 00:34:33.039
But to do this empirical testing in the current environment where for biotech and startup companies like us, we still are dealing with you know financial restrictions.
00:34:33.280 --> 00:34:44.159
It is actually quite complex to figure out a smart development plan because you can't just brute force parallel track all sorts of testing.
00:34:44.320 --> 00:34:54.559
You have to, you know, pick a make decisions along the way and pick a few um educated guesses, I would say, in which direction to go.
00:34:54.800 --> 00:34:59.920
And navigating that from a scientific perspective is complex enough.
00:35:00.000 --> 00:35:22.960
But doing this in a financially restricted manner, um, where you can't do your full developability profile early on is the first step as you're designing a new format, um, purely out of financial reasons, um, is certainly a challenge that I didn't anticipate we would be in for such a long period of time.
00:35:23.039 --> 00:35:44.559
If I look at this at a at an ecosystem perspective, where in general, you know, um the uplifting of that developments on the MA and IPO side in later stage programs still take a while until um, you know, startups like us and biotechs are benefiting from the improvements in the overall financial climate.
00:35:46.239 --> 00:35:47.119
Got it.
00:35:47.519 --> 00:35:47.920
Got it.
00:35:48.000 --> 00:35:49.119
Thank you for sharing.
00:35:49.440 --> 00:35:59.199
Um and now taking a step back a little bit, I want to look more big picture at the state of the industry and some things um that you might be excited by.
00:35:59.360 --> 00:36:14.239
I'm curious to know, you know, looking out at the next couple of years, what developments are are uh exciting you, but also what do you see as the biggest challenges and opportunities in the space that you play in and in maybe in biotech more broadly?
00:36:14.559 --> 00:36:36.000
Yeah, I think um one of the things I'm rattling with a little bit as both an opportunity and a challenge, unfortunately, um is just the general concept of precision medicine, which is certainly where T cell engagers and specifically peptide MHC targeted T cell engagers uh fall into.
00:36:36.159 --> 00:36:40.000
Um that is on the one hand, because of the HLA restriction of our molecule.
00:36:40.079 --> 00:36:42.480
That's one biomarker that you need to consider.
00:36:42.960 --> 00:36:52.320
Um while for some that is already a hurdle, overall, I think this is the one that is that that is the one that we can navigate from a technical perspective.
00:36:52.400 --> 00:37:29.360
Um, but generally the idea of finding the right drug for the right patient to increase the likelihood that a patient will respond because they actually have the target or express whatever uh liability in the tumor you want to leverage, reducing the exposure of individuals who are just not expected to respond because they simply don't have the target or that particular mutation that you're going after, and just improvements in the time and cost to develop the drug if you're not drowning out your clinical signal by enriching for patients that are likely to respond, makes complete scientific sense to me.
00:37:29.440 --> 00:37:40.480
And I'm excited by the opportunities that we have to improve clinical care, even when it remains in an experimental status during clinical trials.
00:37:40.960 --> 00:37:53.280
At the same time, biomarker-selected drugs or clinical development either way, means that you're reducing your patient pool.
00:37:53.519 --> 00:38:26.320
And while I personally find it extremely valuable to have very good outcomes for, let's say, half the number of people rather than poor outcomes for a broader number of people, um, I still feel like there are a few gaps that we need to bridge when it comes to understanding for commercial teams and those who otherwise evaluate the value of a particular drug to really understand why making a difference, even if it is for a potentially smaller patient population, is important.
00:38:26.639 --> 00:38:33.440
So I think that's where a challenge and opportunity still remains in the precision medicine space.
00:38:33.599 --> 00:38:47.760
And more broadly, I guess what I'm excited about as a theme is to realize that more and more companies are starting to think about access of their drug.
00:38:48.320 --> 00:38:59.199
It's not just about making the most innovative, cool, fancy thing that in research hospitals like Dana Farber or MD Anderson can be administered and tested.
00:38:59.440 --> 00:39:09.039
But how do we make drugs that can be ultimately affordable with relatively low cost of goods, which is in part driven by good developability profiles?
00:39:09.119 --> 00:39:28.239
Um, but also learning more about um clinical safety profiles and how to manage adverse events that are often a class effect in community settings are topics that I hear being discussed more and more in companies that are working on solutions.
00:39:28.400 --> 00:39:37.760
I don't think today everything is equitable the way it should be, but it's definitely a topic that will improve clinical cancer care in the long run.
00:39:37.840 --> 00:39:44.639
And I'm excited to know that more and more people are thinking about these topics and then coming up with solutions for this.
00:39:45.199 --> 00:39:46.159
Yeah, me too.
00:39:46.480 --> 00:39:46.880
Me too.
00:39:46.960 --> 00:39:48.880
These are very promising developments.
00:39:49.039 --> 00:39:56.559
And like you said, I think focusing on that accessibility aspect, I mean, that I I hear that as well quite often.
00:39:56.719 --> 00:40:03.599
Um every conference I go to, I think that's that's the underlying theme of improving access, reducing the cost of these therapies.
00:40:03.679 --> 00:40:09.119
And I agree that the development profile, developability profile is a key part of that.
00:40:09.199 --> 00:40:16.880
And so kudos to you and your team for um doing your part and your in your scope of the of the industry to uh to do that.
00:40:17.119 --> 00:40:25.679
So, Johanna, now I want to ask you the question that I ask all my guests on this show, and that is big or small, how do you think we can better biopharma?
00:40:26.480 --> 00:40:28.719
Yeah, I love this question.
00:40:28.880 --> 00:40:31.119
Um, it lets me think big.
00:40:31.360 --> 00:40:44.239
So I get to talk about my favorite topic, which is I wish we could better biopharma by improving collaboration among different companies.
00:40:44.639 --> 00:41:03.760
Specifically, I'm thinking about the translational field where we actually have a lot of patient-derived data sets and characterization of tumor tissues that has been generated and is available.
00:41:04.159 --> 00:41:24.639
However, because we are not sharing these kinds of complex uh data sets, we are hindering our ability to identify biomarkers for particular uh drug, but also in general, in the broader sense, enabling uh new target discovery uh for innovative drugs of the future.
00:41:25.039 --> 00:41:31.280
We can probably spend a whole podcast episode talking about all the reasons why this is so difficult.
00:41:31.440 --> 00:41:40.159
There are obviously hurdles about not people wanting to not share confidential information or intellectual property, which I understand.
00:41:40.400 --> 00:41:47.920
There are technical hurdles about data annotation, sample annotation, even the methods with which data is acquired.
00:41:48.079 --> 00:41:49.360
I get all that.
00:41:49.679 --> 00:42:01.039
I still think it is challenges that are worth thinking about, how to overcome this when it comes to pooling data sets that exist of many, many patients.
00:42:01.280 --> 00:42:04.800
Let's just call it the standard of care comparator arms.
00:42:04.880 --> 00:42:28.960
If we could just have that as a tool available to the ecosystem for better drug development in the future, that would be an immense step forward that also honors the idea that patients selflessly actually allow us to collect biopsies in other specimens and make them available for research purposes.
00:42:29.199 --> 00:42:35.760
So I think the ecosystem would be better for more collaboration on this front.
00:42:35.920 --> 00:42:41.679
Um spanning from discovery to clinical development and care delivery.
00:42:42.960 --> 00:42:46.239
Yeah, I completely, I completely agree with that, Johanna.
00:42:46.559 --> 00:42:47.039
Absolutely.
00:42:47.119 --> 00:42:49.199
And I hope we get to see more of that.
00:42:49.440 --> 00:43:02.320
Um, you know, there's promising developments on the horizon, but I do hope as we move forward, especially in the age of AI, and that data sharing and insight sharing uh continues to grow alongside that.
00:43:02.719 --> 00:43:05.440
Um yeah, it's a great answer to that question.
00:43:05.599 --> 00:43:10.880
And thank you so much for joining me on this episode of Better Biopharma, the official podcast of Bioprocess Online.
00:43:11.039 --> 00:43:13.039
And thank you out there in the audience for tuning in.
00:43:13.199 --> 00:43:14.079
We'll see you next time.
