Latest / Zero Shot Podcast DEMO / Revolutionizing Pancreatic Cancer Monitoring with Liquid Biopsies

Transcript

Revolutionizing Pancreatic Cancer Monitoring with Liquid Biopsies · Zero Shot Podcast DEMO

0:00
Interviewer: Thanks, everybody, for joining OneCell's webinar.
0:04
The topic of today's conversation is advanced liquid
0:12
biopsies.
0:13
And in particular, we'll be discussing not just looking at circulating tumor DNA, but also the emerging
0:21
field of single -cell -based liquid biopsies, which formerly a lot of people referred to as circulating tumor cells, also still
0:29
referred to as circulating tumor cells.
0:30
But we'll also describe what is new about these sort of cell -based liquid biopsies,
0:37
speakers: The five-year survival rate for pancreatic cancer is about 13%. OneCell's OncoInsights technology captures live CTCs for multi-omic analysis. Dr. What claim should we test as this continues?
0:52
Interviewee: So the topic is multisource genomics to monitor clonal evolution for pancreatic cancer.
0:58
Dr.
0:59
Rizal Kurzok from Medical College of Wisconsin will provide an update for the talk.
1:08
speakers: In this conversation, advanced liquid biopsies enhance monitoring of clonal evolution in pancreatic cancer. That opening sets the stakes we'll follow.
1:25
Interviewer: This meeting is being recorded.
1:27
We will have two formats for Q &A.
1:31
As the meeting progresses, please feel free to put your questions in the chat, in the chat box,
1:39
and then we'll accommodate as many of the questions.
1:43
And towards the end, we'll have some time for some live questions as well.
1:47
So the way the agenda is going to work is I'll spend a couple of minutes on just sort of setting the stage, which I'm doing right now.
1:58
This means we're all making patient selection and endpoint decisions based on incomplete, often outdated information.
2:06
But that's a solvable problem, and it's
2:13
speakers: And there's a link there, which you can click.
2:21
CTCs provide a more comprehensive view of tumor heterogeneity compared to ctDNA. Ken Bloom, a strategic advisor to OneCell, will provide insights into their certified laboratory operations.
2:37
Interviewee: I'll turn it over to Ken Bloom.
2:40
Ken is OneCell's advisor.
2:44
He's a strategic advisor to OneCell, also leads our CAPCLEAR certified laboratory in the San Francisco Bay Area.
2:52
Interviewer: He will be providing an overview of OneCell as well as our OncoInsights technology.
3:02
speakers: Pancreatic cancer has a low five-year survival rate of about 13%. OneCell's OncoInsights technology captures live CTCs for analysis.
3:13
Interviewee: next.
3:15
Yes, thanks.
3:16
Thanks, Ken.
3:16
Appreciate that.
3:18
If I can ask all the participants, if you can go to the top of the chat room, there's a link there
3:25
from, it's called forms .gle.
3:29
If you guys just click that link, I promise this won't take more than 10 seconds.
3:33
It's going to ask some basic questions on what are the types of areas that are important to you, and that's it.
3:39
Just click on that, type in your name so that we know who responded to what, and then we can move on to the next part of the agenda.
3:46
So, over to Razelle.
3:50
So, a pleasure to work on this project and with this
3:58
really interesting technology.
4:00
And we really wanted to concentrate on pancreatic ductal adenocarcinoma for several reasons.
4:07
One is because our institution really has a world -class program in this cancer.
4:15
And the second is because it is such a difficult cancer.
4:19
The five -year survival rate is about 13%, despite advances in surgery and systemic therapies.
4:27
And it's actually worse than that.
4:29
There's a recent paper in JAMA that shows that the one -year survival is about 24%.
4:35
So, we have a long way to go to learn how to deal with this cancer.
4:40
And it is projected to become the second leading cause of cancer -related deaths in the United States.
4:48
And so, we really need to improve our understanding of this cancer.
4:58
Interviewer: you whether the target is actually present at the protein level on cells circulating right now during treatment.
5:06
Third is RNA, or whole transcriptome profiling.
5:09
Interviewee: That's about 16 ,000 markers.
5:11
This is your window into gene expression, what pathways are active, what resistance mechanisms might be turning on, and what's changing
5:19
over time.
5:20
Interviewer: For drug programs, this isn't just biomarker discovery.
5:23
This is real -time pharmacodynamic monitoring.
5:26
You can track whether your drug is hitting its target, whether resistance is emerging, and which patients are likely to respond.
5:37
speakers: The implications of CTC genomics for clinical trials and personalized medicine are discussed. Protein expression is measured by immunofluorescence, quantifying drug-targeted proteins like HER2 and TROP2.
5:52
Interviewee: Second is protein expression.
5:55
This is currently measured by immunofluorescence on individual cells.
5:59
You can directly quantify the proteins a drug is targeting, such as HER2, TROP2, clodin -18, nectin -4, etc.
6:07
If you're developing an antibody drug conjugate, for example, this tells
6:16
speakers: OncoInsights captures and analyzes circulating tumor cells, addressing the challenge of their rarity in the bloodstream.
6:26
Interviewee: exactly what OneCells platform addresses.
6:29
Our solution is called OncoInsights, our proprietary platform to capture and comprehensively analyze circulating
6:36
tumor cells.
6:38
Here's a quick refresher.
6:39
When a tumor grows, it sheds cells into the bloodstream.
6:43
These are circulating tumor cells, actual cancer cells floating in a patient's blood draw.
6:48
The problem historically has been catching them.
6:51
They are incredibly rare, maybe dozens or less, among billions of normal cells.
6:56
The trick, though, is not just to capture them, but to capture them in a way that's suitable for downstream analysis.
7:02
speakers: Biopsies predict prognosis and therapy response, but smaller samples limit reliability due to tumor heterogeneity.
7:12
Tumor heterogeneity makes analysis less reliable, akin to understanding a book from a single sentence.
7:20
smaller, our analysis becomes less reliable because of tumor heterogeneity.
7:25
A five -micron section from a one -centimeter tumor represents only a fraction of a percent of that tumor.
7:32
So to put it in a little bit easier terms, that's like trying to understand the entire plot of war and peace by reading one
7:40
sentence of the book.
7:40
Traditional biomarkers have limitations in sensitivity and specificity.
7:48
Interviewer: has emerged as the latest tool in our monitoring arsenal.
7:51
It's real -time, and it reflects what's happening now since the half -life of DNA is on the order of a few hours.
7:58
But CT -DNA only gives you the genetic sequence of the fraction of tumor cells that have recently died.
8:04
speakers: CT-DNA lacks protein expression data and misses emerging clones. CT-DNA misses low-frequency mutations and protein data.
8:14
Interviewee: and of course, it will likely miss emerging clones that are actively proliferating but have not yet died.
8:21
But more importantly, as antibody drug conjugate therapies have emerged, CT -DNA doesn't give us any useful information
8:28
about protein expression or localization.
8:37
speakers: Our labs hold the highest level of clinical accreditation, CLIA and CAP in the U .S., NAVAL and CAP in India.
8:44
And these aren't just logos on the slide.
8:46
They mean that the data coming out of our labs meets the standards required for clinical decision -making.
8:51
Interviewee: Our mission is straightforward but ambitious.
8:54
We're going to focus democratize precision oncology.
8:57
Make it actionable, accessible, affordable for every cancer patient everywhere.
9:03
We are working toward impacting the lives of one million cancer patients through breakthroughs in science and AI.
9:09
For those of you in pharma, you can think of OneCell as a potential partner for clinical trial design, patient selection, biomarker
9:17
discovery, and drug development.
9:19
Everything that you're about to hear is built on that infrastructure.
9:23
So now let's talk about the problem, where we've been and where we're going.
9:27
speakers: Early systems for capturing tumor cells were limited by high noise and low purity, highlighting the need for improved methods.
9:40
The earliest systems utilized single -marker immunocapture and fixed the cells, making them largely unsuitable for comprehensive downstream
9:47
analysis.
9:48
Other systems used negative selection by depleting hematopoietic cells or size -based methods, but these suffered from
9:56
high background noise and low purity.
9:59
OncoInsights solves these limitations with a combination of multi -marker affinity capture and a cleavable linker allowing capture of
10:07
live CTCs with no cell damage.
10:09
Instead of a single hook, we use customized capture beads coated with multiple antibodies targeted to a specific cancer
10:17
type.
10:18
For most cancers, a combination of Epcam and Transferrin works effectively, but the beads can be customized to optimize capture.
10:25
For example, many renal cell cancers don't express Epcam at high levels, expressions quite low.
10:32
So adding CD70 to the beads dramatically increases the CT positivity rate.
10:37
Critically, this allows for the capture of the full diversity of tumor cells, not just the most common ones.
10:44
Once the cells are captured, they are released intact using a cleavable linker system.
10:49
No bead contamination, no cell damage.
10:52
So the cells are ready for multi -omic analysis.
10:56
So this is where things get really exciting because of what you can measure from these capture CTCs.
11:02
The OncoInsights panel covers three biological layers simultaneously from the same blood draw.
11:08
First, DNA.
11:10
1080 gene panel covering point mutations, insertions, deletions, gene fusions, copy number variations, and
11:17
chromosomal rearrangements.
11:19
It also gives you tumor mutational burden, microsatellite instability status, and homologous
11:26
recombination deficiency status.
11:28
All of the genomic biomarkers that predict response to immunotherapy, PARP inhibitors, and targeted agents.
11:39
Interviewee: Any new technology lives or dies on its data, so let's look at the numbers.
11:44
Across more than 2 ,000 patient samples spanning lung, breast, pancreatic, colorectal, liver,
11:52
head and neck, prostate, and gallbladder cancers, the overall CTC positivity rate, or our ability to detect
11:59
CTCs, was 83 % for stages stage 2 through stage 4 patients, but metastatic patients had approached
12:07
100%.
12:07
Let me put that in context.
12:10
Our first generation of CTC capture technology yielded a mean of about 5 CTCs per sample, while our current
12:18
platform yields a mean of about 21 CTCs per sample.
12:21
That's a four -fold improvement, and more cells means more material for downstream analysis, more genomics, more protein
12:29
profiling, more confidence.
12:32
speakers: The limitations of traditional tissue biopsies are discussed, particularly regarding tumor heterogeneity.
12:43
Interviewer: Pancreatic ductal adenocarcinoma progression involves stepwise accumulation of key driver mutations,
12:51
which are mainly known to be KRAS, the cyclin system,
12:59
CDKN2AB loss, and P53.
13:03
And tumor evolution also includes rapid genomic restructuring characterized by
13:10
copy number variations and large chromosomal rearrangements following
13:16
polyploidization.
13:20
So, this is a difficult cancer.
13:24
And the reason, one of the reasons it's difficult is because of the great heterogeneity between patients
13:31
and even within patients with time.
13:34
And that's really difficult to capture if what you're doing is a baseline tissue biopsy.
13:42
Um, so, um, there are a lot of limitations to bulk sequencing.
13:50
Um, and, uh, some of them are illustrated here.
13:55
Uh, so, we use bulk sequencing to understand, um, cancer.
14:02
But the problem is that it doesn't really show us
14:10
An 83 % to 100 % positivity rate across solid tumors means that the vast majority of your enrolled patients can be monitored
14:18
longitudinally with this platform, and that's a game -changer for adaptive trial design.
14:23
One more important point before we move on, and this is particularly relevant for CNS cancers and thoracic malignancies.
14:30
The technology isn't just limited to blood.
14:32
At one cell, we've demonstrated CTC capture, or tumor cell capture, from cerebral spinal fluid
14:40
and pleural fluid.
14:41
In CNS tumors like glioblastoma and medulloblastoma, where a blood draw tells you almost nothing about what's happening in the brain,
14:49
CSF -based CTC analysis opens a completely new window.
14:53
This is significant for clinical trials and CNS malignancies, where biomarker development has lagged far behind other tumors.
15:01
Now, generating all this data is only valuable if you can act on it, and that's what the iCORM platform is designed to do.
15:10
iCORM is one cell's integrated precision oncology consult.
15:14
Think of it as the intelligence layer that sits on top of the laboratory data.
15:18
It has two primary interfaces.
15:21
iDiscover for translational research and clinical trial support, and iCare for the treating oncologist.
15:28
For those of you in pharma, iDiscover is the critical piece.
15:38
speakers: Bulk sequencing cannot distinguish mutational events over time. CTC capture from fluids opens new monitoring windows for CNS tumors. Why single-cell analysis is crucial for accurate monitoring?
15:55
Interviewee: what happens with time, what happens with successive, um, mutational events.
16:02
And, um, it cannot, um, extinguish, it cannot distinguish
16:10
Interviewer: what is happening in different clones of cells.
16:15
So, if we see, do single cell sequencing, we can see, are there co -occurring versus
16:23
mutually exclusive alterations in different cells?
16:27
Might this explain, um, mixed responses to drugs, uh, like PARP inhibitors, KRAS inhibitors,
16:35
and immunotherapy.
16:37
And so, single cell DNA sequencing is really required to resolve the clonal
16:45
architecture of pancreatic ductal adenocarcinoma from early invasion through systemic,
16:52
uh, dissemination and to identify molecular vulnerabilities targetable with precision
17:00
therapies as the, as the disease evolves.
17:03
Uh, so, this is a study that we initiated, um, and, uh, the purpose of
17:11
the study was to isolate single circulating tumor cells from pancreatic cancer patients prior to treatment
17:19
and perform genomic profile to characterize cellular level genomic heterogeneity, and also to
17:26
follow these patients during treatment.
17:29
And, uh, uh, this, uh, study is led by Dr.
17:34
Mondana Kamgar and myself at MCW and Dr.
17:38
Kremolini at, uh, University of Pisa.
17:41
And, uh, we've collected quite a few samples and I think, um, I'm happy to share with you some of the
17:49
insights, uh, from these samples.
17:51
Next, uh, slide, please.
17:55
So, all together, uh, we now have 26 patients, 63 total samples,
18:03
um, and, uh, five time points tracked.
18:07
And, uh, we're looking at CTCs and CT DNA and we're also continuing
18:14
to expand, uh, this, uh, project.
18:18
Uh, but I'm going to show you, I think we have some fascinating insights, um, at this, uh, time.
18:25
And, um, I want to share those with you.
18:28
So, uh, first, um, uh, this slide, I think really illustrates, uh, some of the
18:36
insights, uh, from the study.
18:38
Um, so if we look at, uh, patients that have CT DNA only
18:45
versus, uh, CTC DNA only, we're getting a lot more alterations
18:53
from the CTCs, which you can see on the right, than we are getting from the CT DNA.
18:59
And this in the middle are some of the shared alterations, um, that you see.
19:05
So there's a lot of common things, but themes, but in addition, I think that both the
19:13
CT DNA and the CTCs bring their own information.
19:18
And I think you can see how rich the information is that is coming, um, from the CTCs.
19:25
Um, so this, uh, slide shows, uh, the longitudinal, uh, CTC,
19:33
uh, DNA and, uh, CT DNA longitudinal
19:41
mutational landscape.
19:43
And I'm going to go into, um, more detail in a moment, but if you go to the
19:50
right side, this is patient by patient.
19:53
And if you go to the right side of the slide, you can see, um, on the, um, Y axis on the
20:01
right side, uh, the red is CT DNA and the green is CTCs.
20:08
And we're going to go to higher level magnification, but at, uh, but you can see already from these slides
20:16
that the CTCs and the CT DNA do not always tell you the same thing.
20:23
Um, they give complimentary information sets.
20:28
Um, so this, uh, focuses in, um, on, uh, what we see
20:36
in, um, for instance, CT DNA, which is in the red and CTCs, which is in the green.
20:43
And, um, what you can see is that in some, in some situations we're seeing the same thing,
20:51
but in other situations we actually see, um, uh, more of certain mutations
20:58
in the CT DNA than in the CTCs.
21:03
Um, and, um, so one of the interesting things that we found
21:10
with this is that, um, as we began to look at the CTCs
21:18
versus the CT DNA, um, the, uh, we began to see something that we found
21:25
initially confusing, but now we think it is absolutely fascinating and really goes again,
21:34
uh, along with evidence published very recently in, uh, nature genetics by the Sloan Kettering group.
21:42
Um, so in the CTCs, we're finding, for instance, we find alterations not found
21:50
in the CT DNA.
21:52
In the CT DNA, we find alterations not found in the CTCs.
21:57
And, um, if we put, and some, of course, uh, of the alterations are found both in the CTCs
22:04
and the CT DNA.
22:06
Um, but this is, I'm going to give you an example of something that we thought was really fascinating and actually
22:13
very relevant to therapy.
22:16
Um, so a subset of, uh, patients with KRAS, as the tumor evolves,
22:24
there are CTCs that appear to lose, uh, their KRAS.
22:30
And, um, this was at first a bit confusing to us, but it turns out this
22:38
is very consistent with what other, um, investigators are finding when they're looking at single cell
22:45
populations.
22:46
Um, and, uh, for instance, the Sloan Kettering group in nature genetics about a year ago
22:54
published that about 40 % of the individual clones in pancreatic cancer were
23:01
losing their KRAS.
23:03
And, um, this is precisely
23:10
speakers: AI and machine learning integrate data from CTCs and imaging, enhancing therapy response predictions.
23:18
Interviewee: It integrates multiomic data from CTCs and CT DNA, radiologic imaging, pathology images, and real
23:26
-world evidence data sets.
23:28
It uses AI and machine learning to identify patient populations most likely to respond to a given therapy, flag resistance
23:36
markers as they emerge, and help you understand mechanisms of action in a way that bulk tumor sequencing simply can't provide.
23:44
It also adds the missing temporal layer to AI.
23:48
It allows AI to begin to understand how tumor cells are adapting and what to do about it.
23:54
Imagine being able to enrich your clinical trials for responders before the trial even starts, based on real -time
24:02
CTC protein expression in genomics, rather than on archival tissue.
24:07
That's the promise of iCORM.
24:09
You get smarter enrollment, more informative biomarker endpoints, and ultimately, a cheaper path to approval.
24:17
Interviewer: With that as background, I'll now turn the slides over to Dr.
24:21
Curtis Rock, who's going to present her exciting data on the clonal evolution in pancreatic cancer.
24:26
But first, there will be a few polling questions.
24:29
So, Arshad, I guess that goes to you
24:37
speakers: CTCs reveal alterations not found in CT DNA, reflecting cancer evolution and potential therapy targets.
24:46
Interviewee: what we're seeing as we follow individual CTCs and, um,
24:54
uh, see that there are different clones of CTCs and not all of them retain, for example, um,
25:03
their CTCs.
25:04
So we're, uh, their KRAS.
25:06
So for instance, we're seeing KRAS negative, uh, CTCs.
25:10
Um, sometimes in the KRAS negative CTCs, the MEC pathway is still,
25:18
um, a driver.
25:20
For instance, they may have SMOD4 or they may have other alterations in that pathway, but not always.
25:27
Some of these, uh, CTCs are probably losing different pathways or exploiting different
25:35
pathways in order to, um, propagate, uh, the pancreatic cancer.
25:41
Uh, so this goes back to our prior slide and I want to now focus on
25:49
the Y -axis to the right and, um, um, the green
25:56
part of that access.
25:58
And that's the CTCs.
26:00
And you can see that there's, um, multiple alterations that can be found in CTCs that cannot
26:08
be found in the CT DNA.
26:11
Um, so before we focused on some of the alterations, some of the CTCs that were losing
26:19
alterations, now we're going to focus on, um, some of the alterations found in CTCs that
26:27
cannot be found, um, in the CT DNA.
26:30
And, uh, so you can see there's a whole, um, group of alterations
26:37
that were being, that were finding in the CTCs.
26:42
And, um, these alterations are not found in the bulk, uh, CT DNA.
26:47
And since the CTCs may reflect the evolution of cancer, pancreatic cancer, and other
26:55
cancers, and also may be seeding some of the metastatic sites, these additional alterations
27:03
may be critical to thinking about therapy.
27:10
speakers: Tumor heterogeneity worsens over time, with new alterations appearing, highlighting the need for comprehensive monitoring.
27:19
Interviewee: we have clear cut evidence for it, um, pancreatic cancer and probably every other cancer has a
27:27
lot of tumor heterogeneity.
27:29
And in addition with time that tumor heterogeneity, um, becomes
27:37
worse.
27:38
It undergoes molecular evolution and, uh, there are, uh, new alterations that are appearing.
27:45
There may be old alterations that are muted or disappear.
27:54
And, um, to reiterate tissue biopsy or just bulk CT DNA
28:02
cannot outline all that is happening at a single cell level.
28:07
And so CTCs demonstrate molecular diversity of the tumor at the baseline, uh, which is
28:15
hidden with bulk CT DNA and tissue analysis.
28:24
Interviewer: And then, um, the final, um, part of this, uh, slide, if you look on the, on
28:32
the Y axis to the right, uh, the bottom of the right axis, you can see the red dots.
28:38
And this is the CT DNA only.
28:41
So there are alterations also found in CT DNA that are not found in CTCs, um,
28:48
less so than vice versa, but there are certain alterations in some of our patients.
28:54
Uh, this shows, um, some of the alterations found in individual patients in the CT DNA and not
29:02
found, um, um, in the CTCs.
29:05
Interviewee: So when we put all of this together, the bottom line is we need both sources of information.
29:12
We need the CT DNA and we need the CTCs.
29:16
And this again puts together, um, some of the alterations found in the CT DNA only, some of the
29:24
alterations found in the CTCs.
29:28
And, uh, there's a lot of alterations that are found in the CTCs that are not found in the CT DNA and then some of
29:36
the shared alterations.
29:38
So these are complementary techniques and this type of liquid biopsy is really unique in that it can
29:46
give, um, investigators the full picture of what is going on in an individual patient.
29:53
And, um, likely the difficulty with pancreatic cancer, metastatic pancreatic cancer, and
30:01
probably many other, uh, cancers is that we've been getting a very partial picture.
30:08
Um, if we do tissue biopsy, we get a picture at one point in time of one small piece of tissue.
30:16
If we do a liquid biopsy, a blood biopsy, but we do CT DNA, we can get a serial picture
30:24
of what is happening in the cancer, but it's a bulk serial picture.
30:29
And we're not seeing what is happening in individual cells, which may reflect a tremendously
30:37
important, um, biology of the cancer and also may reflect, um,
30:44
seeding of new metastatic sites
30:46
Interviewer: and how those clones may be different, um, may be the key, uh, to understanding, um, our
30:54
next, uh, step in steps in therapy.
30:58
So, um, this, um, emphasizes the point that we
31:05
kind of always known, but here
31:15
CTCs allow you to follow that on a regular basis and see what is
31:22
happening at a single cell level as the cancer progresses.
31:27
So, uh, they're important at baseline and they're important through the life cycle of that, uh, cancer.
31:35
Um, so single cell CTC genomics provides biologic insights beyond
31:42
conventional, um, liquid biopsies and it, uh, reveals tumor subpopulations and
31:50
genomic alterations not captured by CT DNA, broadening the detectable molecular landscape and metastatic
31:58
pancreatic cancer beyond what circulating DNA alone can resolve.
32:03
And again, we're talking about pancreatic cancer, but it's very likely, and that's what our project was on, but
32:11
it's very likely that this really applies to all tumors that tend to metastasize.
32:17
Um, so I want to loop back to this really interesting finding,
32:24
um, that our, uh, project has confirmed prior results, um, from,
32:32
uh, uh, that were published in nature genetics about a year ago.
32:36
Um, so, um, to our surprise and probably to the Sloan Kettering group surprise, uh, when they
32:44
published this, um, we identified CTCs that have KRAS, that are KRAS independent.
32:52
They appear, this is one example, but they appear to have lost their KRAS, um, and they're circulating
33:00
in the bloodstream.
33:01
Um, so this would be invisible to CT DNA based assays, um, that, uh,
33:10
track, uh, tumor DNA in bulk.
33:14
Interestingly, in these KRAS independent clones, in some of them, alternate pathways are
33:22
activated in the MEC signal, MAPK, SMOD.
33:28
Um, and in some cases, um, and in some cases, just completely alternate pathways are activated.
33:35
PI3 kinase, AKT, uh, genomic instability, JAK -STAT.
33:41
And I think we really have to have a good understanding of exactly
33:50
Interviewee: Um, but it's, appears that patients have multiple, each patient has multiple
33:58
different clones that may vary from each other.
34:03
And, um, CTCs enable this high resolution delineation of
34:10
coexisting clonal population and the genomic underpinnings of, uh, what is probably
34:18
emerging resistance mechanisms when we treat.
34:22
Um, so, um, uh, looking at both CT DNA and CTCs,
34:30
it enables dynamic longitudinal tracking.
34:35
speakers: The integration of CTC analysis into clinical practice represents a significant advancement in personalized medicine for cancer treatment.
34:46
Interviewee: Serial single cell CTC analysis enables real -time monitoring of clonal evolution during
34:53
therapy, and it offers a direct window into tumor adaptation, uh, selective pressure responses,
35:01
and treatment -driven genomic shifts over time.
37:20
speakers: Now other alterations are emerging, which may be very important, including BRCA, uh,
37:26
RAD51L, and, um, MUTYH.
37:32
Um, a couple of these alterations like BRCA and RAD51 may be causing HRD
37:39
and, uh, could be, uh, targeted with, um, PARP inhibitors or platinums.
37:46
And then with more time, uh, the BRCA is retained and the P53 is retained.
37:53
Um, but a P53CA alteration emerges.
37:57
And so I think, um, this is really, um, important for understanding what's happening in the
38:05
patient and how is resistance emerging and how should we mold our
38:12
therapies, um, to this evolution that is occurring in an individual patient.
38:20
And that is really what the ultimate goal is of precision or personalized medicine is to understand
38:28
each patient in detail at each point in time of their disease.
38:33
So we can mold the therapy that it would be best suited to that patient rather than guessing at what the
38:41
next therapy should be.
38:43
Um, I think this is really next generation precision oncology.
38:48
Um, and, uh, this allows us to break the resolution barrier in precision oncology, allows us to understand
38:56
biologic, uh, resolution, resistance detection, uh, tumor heterogeneity, um,
39:03
all the different mutations and alterations.
39:06
Um, and, um, um, um, we can go from population average genomics to single
39:14
cell tumor biology, single cell genomics.
39:18
Um, and in addition, uh, CTCs allow one to go beyond DNA.
39:24
And this is really important.
39:26
Um, CTCs allow one to look at, uh, DNA, but it, one can look at RNA.
39:32
One can look at proteins on the same cell.
40:06
Rizal, is there any other slide you wanted to discuss in terms of alternative pathways?
40:11
No, that's it.
40:11
I'll just say, okay, perfect.
40:12
Thank you, Rizal.
40:13
Really appreciate, uh, the presentation.
40:16
Uh, we already have some questions coming in, but just to remind you guys, as you can see on the slide, uh, we will be at
40:23
AACR and I know most of you are likely to be at AACR.
40:28
So would, if there are any questions or if you would like to meet us, you can easily, uh, take a picture of our, put your camera in front of this
40:36
QR code and schedule a meeting with us.
40:39
It could be 15 minutes, it could be 30 minutes, but it'll be great to see you guys.
40:44
Okay.
40:45
Okay.
40:45
Let's move to the Q and A part.
40:47
Um, I see some questions coming in already.
40:50
Um, excellent.
40:52
So we have, okay.
40:54
So Sejong Chun, uh, asked a question on, yeah.
40:58
So what's the, what's the difference you think, Rizal between the difference that why does CTCs, do you
41:06
have a thesis on why they pick up unique and more mutations in ctDNA?
41:13
Um, why do CTCs, well, I think your, um, one reason you may be picking up, uh, more mutations than
41:20
ctDNA is because, um, uh, you're looking at multiple different CTCs and you have different clones.
41:27
And, um, so, uh, that may be a reason to pick up, uh, more mutations than ctDNA.
42:25
Um, uh, so I want the technical person to answer it, but I'm going to tell you what I think.
42:30
Um, I think there would still be a difference, uh, because there may be, um, uh, a, a, uh,
42:39
small numbers of CTCs that have a difference and that technically can't be picked up in the bulk
42:46
ctDNA.
42:47
And yet, even though it's small numbers of CTCs that have that difference, those CTCs may be critical to, um,
42:55
what is happening in the patients with, um, seeding of new metastatic sites or, or reflecting resistant clones.
43:03
So that's my answer, but somebody who's more technical than me might, um, have, uh, be able to add to that.
43:11
Sure.
43:11
I'll chime in on that one.
43:13
Uh, there, there's definitely differences, but before I get into that, I'd just like to add one other point.
43:19
If you draw CtDNA at, you know, draw, take a sample, send it to your favorite CtD, you know,
43:27
garden, your favorite CtDNA lab.
43:29
Draw it in the afternoon, send it to Tempest, draw it in the evening, send it someplace else.
43:33
You won't get the same mutations back, right?
43:36
You're going to get, you're getting basically a small window of cells that are dying at that point of time.
43:41
Now, your question I think was more to the sensitivity.
43:43
And if I raise the sensitivity of the assay, will I be able to detect things?
43:48
And the problem with assays is generally when you raise the sensitivity, you lower the specificity.
43:53
So now I start picking up, you know, it, where do you want to live and die, right?
43:57
You want to get more false positives, more false negatives.
44:00
But in the pharma's perspective, that might be slightly, um, not the way
44:04
Interviewer: think.
44:05
So we would like to see the drivers or the targetable ones, which would probably, um, consist of the, um, the larger proportion of the
44:13
mutations detected, which would be relatively consistent across different vendors and, um, different time points as well.
44:20
So that, um, that, that's, that's a conundrum that we're facing all the time.
44:24
But, um, this sort of leads to my next question where with the CTCs across multiple different time points, there were a few mutations that were
44:31
consistent, but there were other ones that appeared and then went away.
44:34
So I'm a little curious on how those were calculated.
44:37
Um, do you think they're drivers of, let's say resistance because, um, in, in that patient, you have the biomarkers going down and eventually
44:45
down the line as tumor burden starts to increase, you see more mutations happening, but at the end of the day, it did seem that TP53
44:53
was the, let's say, surviving clone later.
45:00
I, uh, do you want me to try to answer that?
45:03
Yeah, Ray, that's for you.
45:05
Interviewee: Yep.
45:05
Um, yeah.
45:07
So we actually found that fascinating and, um, uh, uh, we, uh, especially
45:14
found, uh, yes.
45:16
So P53 was the surviving clone in that patient, but we, with time, we don't know what else will come up, but what we
45:24
found most fascinating, I think was the loss of some of those clones, like the loss of K -RAS.
45:32
And, um, uh, to be frank, uh, that totally surprised me, but it was completely
45:39
consistent with the Sloan -Kettering Nature Genetics paper, which, um, I think also surprised us.
45:46
And the reason I think that's really important is from, to me, I'm, um, precision medical
45:53
oncologist, is because if we're looking at therapies, if, um, the clones change, this may
46:01
explain, uh, resistance.
46:03
So if the remaining clones, uh, no longer have K -RAS, that may explain, um, I mean, I
46:11
had all sorts of other explanations before some of this data came out for why patients might be resistant.
46:17
And I think some of those explanations still hold that there may be co -drivers, but now we can add to that explanation
46:25
that there may be loss of certain clones.
46:27
We may have actually killed off certain clones.
46:30
Um, and that means we have to adjust our therapy, um, accordingly.
46:39
Thank you.
46:40
All right.
46:40
Excellent.
46:42
Um, Dean, uh, you had a question.
46:44
I can ask it, but it's better coming directly from you.
46:47
I don't need to put you on the spot.
46:49
Are you still there, Dean?
46:51
Yeah, I'm here.
46:53
Go for it.
46:54
Yeah.
46:54
So, uh, how often are the, how often do you see consistency between the clonal profiles and
47:02
frequencies or profiles say at day one versus after targeted therapy?
47:08
And, and do they get killed in a way that's consistent with the mechanism of the targeted therapy?
47:19
Interviewer: Uh, is that for me?
47:21
Sure.
47:22
Yeah.
47:23
Um, so I don't think we know that answer completely.
47:26
Um, uh, this is why we're doing the study.
47:29
Um, uh, so, uh, in order to determine, so I, I can't answer that.
47:35
Um, I, I, I just don't think we have enough data to know.
47:39
We have done other prior studies, uh, that suggest, um, that, uh,
47:48
the clones do disappear consistent with the, uh, mechanism of the targeted therapy and the emerging
47:55
clones, um, have different alterations or, uh, new mutations.
48:01
Uh, but this is part of this study trying to figure out, um, is this consistent with the therapy that
48:09
we are giving?
48:11
So I don't think
48:11
Interviewee: have the data yet.
48:13
Okay.
48:14
Thank you.
48:14
And the follow up question was around, um, you know, it's 10, it looked like the early graph said between 10 and 50 CTCs
48:22
per patient per draw.
48:24
Interviewer: Um, is, could you do some statistics on those?
48:28
I, I see they only reported, you only reported presence or absence of a mutation in each sample, but if you had 10 or
48:36
20, you could say 15 % were, had the mutation.
48:41
Interviewee: Have you looked into that?
48:43
We haven't looked into it, but it's a good idea.
48:46
Okay.
48:47
Um, but I still think, um, if there's substantial numbers of cells that are losing the mutation,
48:55
that's telling you something important, but the actual percentage would be good to look at.
49:00
We agree.
49:02
Thanks.
49:05
All right.
49:06
Feel, guys feel free to either raise your hand and ask a question directly or add it to the, to the chat box.
49:13
Um, Ray, I had a question for you.
49:16
Um, is there, uh, so you've shown the differences and the overlaps, right?
49:22
In terms of number of mutation and the types and the, and the actual mutation.
49:27
Did you observe any differences in the types of mutation between CTD and a versus CTCs?
49:33
Miss sense versus nonsense versus so on and so forth.
50:05
speakers: Um, uh, maybe, um, the one cell people have more information in that regard.
50:12
Um, but, um, I, I, I think that there's definitely differences in the
50:19
types of mutations that we're picking up.
50:22
There's also convergence in the types of mutations, um, that we see on CTD and a and CTCs, but there's differences.
50:30
And one of the things that I found really interesting was in, uh, some of the CTCs, uh, maybe
50:38
there would be loss loss of KRAS, but there would be other, um, mutations that activate the MEC pathway,
50:46
uh, that would emerge, um, at the same time.
50:50
Sometimes those other pathways didn't emerge, but sometimes they do emerge other mutations that keep that pathway going.
50:58
Interviewer: Yeah.
50:58
Yeah.
50:58
Arshad, just to add on a, just a quick comment.
51:02
That this is one of the reasons why having an AI platform supporting all this is so important.
51:08
You know, we're, we're in brand new territory.
51:10
All these are fantastic questions.
51:12
We don't know the answers to the most of them because the data is just being generated now.
51:17
And I think it's also highlighting how difficult it is to, to understand what these all really mean.
51:24
We have to come up with a hypothesis and then we just look and say, does it support it or not support it?
51:29
But there's a lot of data hidden in all of this.
51:33
Great.
51:34
Uh, and so this is a question to both you and Ken, um, put yourself in the shoes of a, a, someone
51:41
leading a clinical trial, um, either for a targeted therapy or for an ADC.
51:47
I know, Ray, you talked about ADCs as well.
51:49
Um, how, what's the implications for them in terms of how is this approach of
51:57
advanced liquid biopsies useful to them from a clinical development perspective?
52:24
speakers: Um, the second is, uh, with time, um, if we look at what is
52:32
happening with the CTC evolution, um, it can probably tell us early
52:39
on, um, if the patient's going to respond.
52:44
Um, so we haven't yet done, um, these studies, um, here, uh, but we know from
52:52
other studies, uh, that we can tell, uh, response with liquid biopsy, uh, way before,
52:59
uh, two months, um, with a scan.
53:03
And that, um, is true with targeted therapies is true with chemotherapy.
53:08
So there may be early indicators of response, which we have not had the opportunity yet to, um, show
53:16
here, but I think that, um, there's a lot, there are several studies, some of which we have done that show, um, that, um,
53:24
looking at these serial liquid biopsies, we can see early indicators of response.
53:29
Um, third, um, it may tell us, um, early indicators of resistance, um,
53:37
and not only which patients may be resistant, but why they may be resistant.
53:44
Uh, so it may be because the CT DNA increases.
53:49
It may be because the CTCs increase, but there may be resistance even with a
53:56
decrease in CTCs, if the CTC clones no longer have the target.
54:02
Uh, so I, I think this allows us to decipher, um, uh, the
54:09
heterogeneity in response and resistance, um, that, um, is probably more
54:17
complicated and more interesting and more informative, um, than we previously thought.
54:25
And certainly gives us a lot more information than doing a scan every two or three months or, uh, just looking at CT
54:33
DNA.
54:38
Interviewee: Great.
54:39
Thank you.
54:39
Ken, was there any follow up on that?
54:40
Yeah.
54:42
I'll just add from a pathologist perspective, certainly with ADCs, a lot of the patient selection is based off of biopsy
54:50
material, some of which can be actually quite old.
54:53
Uh, sometimes you can't even find the biopsy on the patient.
54:57
Uh, but you should know anytime you take a biopsy and it's fixed in formalin and prepared for, you know,
55:05
immunohistochemistry, those epitopes have been severely altered by the pre -analytics of, uh, of getting the sample to begin with the
55:13
ischemic times, et cetera, fixation, cross links, proteins, and all sorts of bizarre ways.
55:18
Interviewer: We have to do retrieval mechanisms to try to bring those epitopes back.
55:23
Sometimes we bring them back.
55:25
Well, sometimes we open up, uh, to nonspecific findings.
55:29
And so the ability to be able to capture a CTC at the time point of therapy and probe it for proteins
55:37
without any of those effects.
55:39
And I think maybe that wasn't really clear.
55:41
When you capture a live CTC, you don't have to worry about ischemic time fixations, antigen retrieval, all of the things that
55:48
are negatives for immunohistochemistry, wind up going away with CTCs.
55:54
And most importantly, when it's intact, we can absolutely identify and quantify the amount of protein on the surface of the
56:01
cell that their drug would be, uh, available for rather than relying on pathologist eyeballs or AI techniques to
56:09
delineate membrane from cytoplasm.
56:12
I think there's a lot of advantages now.
56:14
Everything else I agree totally with Dr.
56:16
Kershra said, I think she eloquently pointed out all the advantages.
56:20
Interviewee: All right.
56:20
We have two more questions.
56:21
I'm going to quickly go through those.
56:22
So Langdon Miller asked, uh, that one of the critical challenges, uh, is when you're measuring resistance as baseline is what are the
56:30
intrinsic, uh, mechanism resistance or sensitivity mechanisms that may steer the therapy towards one patient or another.
56:37
Are there any known genetic profile in pancreatic cancer that predicts sensitivity or resistance to agents that are commonly employed to treat the
56:45
disease?
56:46
For example, taxanes, gemcitabine, 5 -FU, oxapleton, illinotican, et cetera.
56:51
Interviewer: Uh, so, uh, okay.
56:56
So, uh, in pancreatic cancer in particular, and with chemotherapy in particular, I think that, um,
57:04
uh, we don't know profiles very well, um, and what predicts, um, but I think we're learning and I'm going
57:12
to give you, um, just a couple of examples.
57:14
Um, uh, profile, uh, cells that have, uh, homologous repair deficiency are likely to
57:22
respond to platinums or, uh, to PARP inhibitors.
57:28
Um, we have a paper, uh, out that, uh, shows that, uh,
57:36
uh, uh, profiles that include ATM, um, or check to alterations, um, which
57:44
are usually thought of HRD related genes, but they're actually, um, they can cause HRD, but mostly they don't,
57:52
uh, but they sensitize to topo one, uh, inhibitors, topo one isomerase inhibitors, such
58:00
as irinotekin.
58:00
Um, as far as gemcitabine and taxane, I think that there's a lot less known about what
58:08
profiles, um, might sensitize to those chemotherapies.
58:12
But I will also say that I think the next generation of therapies, uh, for pancreatic
58:21
cancer probably needs to be better than gemcitabine and, uh, abroxane.
58:26
Um, I, I think there's a lot of new drugs coming out and this can inform our use of next generation drugs.
58:33
Okay.
58:34
Before we end, one last question.
58:36
Apparent loss of KRAS, is that supported by orthogonal biopsy methods such as IHC or histology?
58:44
Um, we, uh, yeah, so that's a really good question.
58:48
Um, so we are now cross -checking this loss of KRAS with, um, um, really looking at the cells, making
58:55
Interviewee: sure that, um, uh, that the IHC profiles are correct for pancreatic cancer and so forth.
59:03
Um, but, um, the bottom line is that I feel very confident that this is a real result.
59:09
And the reason I feel confident that this is a real result is first of all, because the technology is really good.
59:15
And second of all, um, we, we are basically, uh, confirming that this technology shows
59:23
exactly what Sloan Kettering, uh, showed when they looked at single cell, uh, technology.
59:28
Uh, when you look at this in bulk, you don't see the loss, but when you look at single cell technology, um, in the nature,
59:36
uh, genetics paper, it's a paper by Zhang, Z -H -A -N -G.
59:40
If you want to look it up and it was published at the end of 2024, they found that 42 % of, uh, the clones were
59:48
losing KRAS, which I think is a real surprise.
59:51
But the single cell technology gives
59:53
Interviewer: you that and bulk technology cannot.
59:57
Great.
59:57
Thank you.
59:57
That was from, uh, Amar Adam.
59:59
So thank you, Amar, for that question.
1:00:00
With that, uh, really appreciate Ray, uh, for taking your time to, to speak to us and sharing your, your insights.
1:00:08
speakers: Uh, Ken, thank you as well.
1:00:10
And I really appreciate all of the participants.
1:00:13
Uh, and we will look forward to speaking to you and seeing you, uh, at, uh, AACR.
1:00:18
All right.
1:00:19
Thank you.