IsoPlexis Roundtable/Literature Review - 4

EXPERT PANEL DISCUSSION
changes almost as rapidly as the spectrum of new
drugs becomes available. And it is all a feedback. We
would not be making these drugs if it was not for the
types of measurements that we have the capabilities to
do now.
Dr. Merghoub: Not only is it context dependent, but
also we need to actually target multiple drivers of an
immune response. For example, interferon-gamma response signature, PD-L1 expression, presence of T cell
infiltration, looking at immune-suppressive mechanisms, such as presence of Treg, myelosuppressive
cells, IDO, etc.
Establishing some sort of immunogram of each
patient telling you what the relative contribution of
these is, depending on the context and having some
kind of ratio threshold of effector versus immunesuppressive mechanism, would be a good way to do it.
Of course, the ultimate goal is to have means to do it
less invasively. The best way, right now, is to look in
the tumor. But if we interrogate both tumor and periphery or whatever is accessible, we will be learning
more and more what correlates with what we see in the
tumor.
Dr. Butterfield: One thing we have not mentioned yet
in terms of types of biomarkers is microbiome. We
realize that it is critically important and that it could
help mediate resistance. However, we have lists of
bugs, and what we probably need is function of those
bugs and what the function of the population of bugs
is. It is another important biomarker that we have not
mentioned yet.
Dr. Heath: I will strongly second that. There were
indications over the past couple of years that at some
level, one can actually predict the microbiome from
metabolome and vice versa. It is not really clean yet,
but it is there. But to do really targeted metabolomics
and identify metabolic susceptibilities is challenging,
much less at the single-cell level, being able to do such
targeted assays. And we know they are present, because if you do find them, you can find that case where
they work.
Dr. Daver: The molecular field has taught us a lot.
Initially we thought it was going to be simple: you
have a particular mutation, you use targeted therapy,
and it is going to work. In some subsets, it worked
well, and in some subsets it did not work. And in many,
even after it worked, you would see relapses occurring
frequently.
And so now we have started, at least in leukemia,
where we have a lot of data on different molecular
annotations, prognosis, and therapy, that it is not just
one mutation, but it is the comutations (maybe a
4

second and third mutation) that can frequently further impact the overall prognosis and response to
therapy.
Unfortunately, this is not going to be a simple answer. I think we are going to start getting large-scale
immune profiling, which will include different cell
subsets, different functional parameters, and then
hopefully get us close to 80%, 90%, and 95% predictability. And that is kind of how things are moving
in the molecular field now.
Dr. Merghoub: I like how you make a parallel between the targeted therapy world and immune therapy,
because I feel that this is lacking, right? Probably only
people who have experience in the immunotherapy
space think more of personalized medicine.
But in general, it is catching up. It is not there yet.
And I think it is important to give that notion and
convince people that we cannot treat every patient the
same way, and that having a grasp on what each patient's specific profile is important to design the therapy that is appropriate.
Dr. LeMieux: How do single-cell solutions address
gaps in the current biomarker offerings?
Dr. Butterfield: It is very critical. Having an average
of what a group of cells is doing in part of the tumor
gives you a first-level insight. But without understanding what the program within an individual cell is,
you cannot really understand what is going on in different places in the tumor and trafficking of these cells.
This has been another wonderful revolution in biomarker research.
Dr. Merghoub: One of the unique characteristics of
single-cell analysis is the fact that the characterization
is not biased by a preselection of markers that we
would have predetermined.
We are not saying, ''What is happening when you
look at 20 markers that I have a predesigned base of
prior knowledge?'' And that is really the strength of
this approach, and it has allowed us to see things that
we were not able to predict.
The only problem with these techniques is that it is
very onerous, and has to be done in real time, generally. So, there are quite a few limitations there. In
general, now, we only look at *10,000 events, which I
think is quite limited. But it would be nice if that
evolved quite a bit.
Dr. Heath: We have done a lot of single-cell stuff in
my laboratory that has had a transformative effect in
how we think about tumor biology. And in fact, it has
elevated these single-cell methods, and the associated
new emerging immunotherapeutics have really made
ª 2020 by Mary Ann Liebert, Inc.

IsoPlexis Roundtable/Literature Review

Table of Contents for the Digital Edition of IsoPlexis Roundtable/Literature Review