Cellular determinants of anti-tumor immunity in human cancers

Transcript

• 00:00To do is just briefly
• 00:02go over kind of our
• 00:03approach towards
• 00:05studying cancer immunology in humans.
• 00:07And my particular interest moving
• 00:09forward is going to be
• 00:10in studying immune and cancer
• 00:12cell interactions and how
• 00:14they shape
• 00:15anti tumor immunity overall.
• 00:18But we do this in
• 00:21by by first taking a
• 00:22look at
• 00:24patient specimens, clinical questions, relevant
• 00:27clinical questions,
• 00:28and then using
• 00:29techniques to try and understand
• 00:31fundamental aspects of human immunology.
• 00:33And so to do this,
• 00:34we first have simple questions
• 00:36using high dimensional techniques such
• 00:38as single cell sequencing.
• 00:41For example, what who are
• 00:42we looking at? What is
• 00:43the transcriptional,
• 00:45heterogeneity that is present within
• 00:47tumors and immune cells?
• 00:49What is the function and
• 00:50what is the interaction that's
• 00:52going? And in the case
• 00:53of t cells, which is
• 00:54where our biases
• 00:56lies, we then ask, you
• 00:57know, where else can we
• 00:58trace these t cells to?
• 01:00And,
• 01:02how are the microenvironment really
• 01:04influencing,
• 01:05Timur and Mindy? And, ultimately,
• 01:08with
• 01:09the the phenotypes that we
• 01:10identify and
• 01:12these fundamental
• 01:13mechanisms that we identify, how
• 01:15can we leverage this back
• 01:16into the clinic for clinical
• 01:18intervention?
• 01:19And so, one vignette I'd
• 01:21like to just bring up
• 01:22that's published that I'll briefly
• 01:23go over, and this is
• 01:24actually, a project I was
• 01:26co mentored with doctor Kluger
• 01:27and funded by the Skinspor,
• 01:30that we did in the
• 01:31extracranial
• 01:32setting with patients with melanoma.
• 01:34We first identified
• 01:35that
• 01:36there's
• 01:38a
• 01:39subpopulation
• 01:40of CD8 T cells that
• 01:41is tumor antigen specific,
• 01:44but is actually highly resembles
• 01:47recently described,
• 01:48CD8 regulatory T cells, and
• 01:49so it dampens antitumor immunity.
• 01:53We were able to trace
• 01:54these out into the blood
• 01:56and found that the levels
• 01:57that we find in the
• 01:58blood correlate with those that
• 01:59are in the tumor.
• 02:01And as mentioned, they impair
• 02:03antitumor immunity, and what we've
• 02:05found was that it does
• 02:06so by targeting other antigen
• 02:08specific t cells
• 02:10and that ultimately this results
• 02:12in
• 02:13poor patient outcomes.
• 02:16But the second vignette that
• 02:17I'd like to focus on
• 02:18really stems from our ability
• 02:19to leverage kind of
• 02:22fundamental aspects that we've learned
• 02:24through the years
• 02:26and to bring them back
• 02:27into the clinic. And so,
• 02:28this is ongoing work that,
• 02:30is unpublished,
• 02:31but,
• 02:32that
• 02:34we are working
• 02:35on right now in patients
• 02:37with glioblastoma.
• 02:38So
• 02:39in patients with glioblastoma,
• 02:40immune checkpoints really have not
• 02:42worked well. Anti p d
• 02:44one regimens have failed to
• 02:45improve survival for patients,
• 02:47in multiple
• 02:48phase three trials. Hypothesized mechanisms
• 02:51of failure include
• 02:53activation of regulatory t cells,
• 02:54so the suppressive t cells,
• 02:56in addition to changes in
• 02:58the cancer cell plasticity and
• 03:00glioblastoma cancer cells,
• 03:02among many other regimens. But,
• 03:04a couple years back, we
• 03:06leveraged,
• 03:07this kind of institutional expertise
• 03:09in regulatory t cell biology
• 03:10and identified,
• 03:12alternative,
• 03:14coin receptor called TIGIT,
• 03:16which we understand to be,
• 03:18unlike PD one, more important
• 03:20for stabilizing the suppressive function
• 03:23of regulatory T cells.
• 03:24This is work that was
• 03:26done out of the Haffler
• 03:27lab that suggests that,
• 03:29TIGIT expressing regulatory T cells
• 03:31are more suppressive than TIGIT
• 03:33negative, not nonsuppressing T cells,
• 03:35and that when placed into
• 03:36pro inflammatory environments,
• 03:38TIGIT engagement, TIGIT signaling
• 03:41through CD one five five,
• 03:43helps stabilize the suppressive phenotype.
• 03:46Now we also know that,
• 03:49TIGIT is highly expressed in
• 03:50addition to its binding partners,
• 03:51highly expressed in glioblastoma,
• 03:55unlike in multiple sclerosis, which
• 03:57is a pro inflammatory
• 03:59condition in the CNS.
• 04:02And that therapeutically targeting both
• 04:04TIGA and PD-one
• 04:05in humans and also
• 04:08in preclinical models,
• 04:10helps improve pro inflammatory
• 04:12conditions.
• 04:13And so,
• 04:14this led to the institutional
• 04:16initiated trial,
• 04:17led by doctor Amearl and
• 04:19close collaboration with doctor Malinturno,
• 04:22and since taken over by
• 04:24doctor Kurz,
• 04:25for this IIT that, is
• 04:26investigating the combination of anti
• 04:28PD one and anti TIGIT
• 04:30in patients with recurrent glioblastoma.
• 04:32And this is a two
• 04:33part trial. The first part
• 04:34was a safety lead in.
• 04:35But the second part, right,
• 04:37is a perioperative condition where
• 04:39patients get
• 04:40one of four treatments prior
• 04:42to going to surgery, so
• 04:44either antitigid alone, anti PD-one
• 04:46alone, the combination, or placebo
• 04:48followed by combination afterwards. And
• 04:50this really allows for us
• 04:52to collect
• 04:53tissue
• 04:54specimens to try and understand
• 04:56aspects of
• 04:57tumor immunity.
• 04:59And so we've designed
• 05:02translational studies to try and
• 05:04really
• 05:05deeply
• 05:06analyze what
• 05:08are the perturbations that occur
• 05:10following these perioperative
• 05:12treatments.
• 05:13And, the the trial has
• 05:14just
• 05:15concluded
• 05:17accrual, so a lot of
• 05:18these studies are ongoing. Has
• 05:28referenced earlier is being run-in
• 05:30close collaboration with Rong Fan
• 05:31and Yang Liu, in addition
• 05:33to Marcello Distasio, who's a
• 05:34neuropathologist.
• 05:36But we do have some
• 05:37preliminary data that does give
• 05:39us, some optimism that, you
• 05:41know, biological activity is occurring
• 05:43in these patients.
• 05:45And these include that at
• 05:46very early time points, we've
• 05:48observed a
• 05:49shift towards effector phenotypes,
• 05:51effector t cell phenotypes in
• 05:53circulating populations in the blood.
• 05:55And so what I'm showing
• 05:56on the bottom are results
• 05:57from flow cytometry data from
• 05:59our safety reading
• 06:00that suggests at day five,
• 06:02as early as day five,
• 06:03we see a shift from,
• 06:06more naive,
• 06:07or memory populations towards effector
• 06:09populations.
• 06:10This is also reflected in
• 06:12their ability to secrete cytotoxic
• 06:14th one cytokines.
• 06:15This occurs from from both
• 06:16CD8s and CD4s.
• 06:18And based off of our
• 06:19preclinical,
• 06:21hypotheses,
• 06:21TIGIT would,
• 06:23differentially impact,
• 06:25regulatory T cell phenotype. We
• 06:27also do see shifts in
• 06:29the phenotype,
• 06:30down regulation of FOXP3 and
• 06:32up regulation of regulation of
• 06:32CD226 in circling Tregs as
• 06:35well.
• 06:36And, I guess, not shown
• 06:38here is, we also see
• 06:40a
• 06:41relative trend towards decreased suppressive
• 06:43function in circling Tregs.
• 06:46In one patient who we
• 06:48were able to collect both
• 06:49pre and post treatment,
• 06:52tumor samples,
• 06:53we ran single cell RNA
• 06:54and T cell receptor sequencing.
• 06:56And what we observed was
• 06:57that, at the post combination
• 07:00treatments
• 07:01sample, we do see an
• 07:02infiltration of,
• 07:03the absolute quantity of,
• 07:06infiltrating T cells as represented
• 07:08by CD three staining,
• 07:10but also a shift in
• 07:11the the amino phenotype. So
• 07:13we see a much larger
• 07:14expansion of
• 07:16effector
• 07:17CDT cells,
• 07:19in addition to actual clonal
• 07:20expansion.
• 07:22But maybe most interesting to
• 07:24us based off of our
• 07:25preclinical observations was that if
• 07:27we were to take the
• 07:28t cell receptor sequence and
• 07:29use it as a molecular
• 07:30barcode
• 07:31and track what is the
• 07:32phenotype of regulatory t cells,
• 07:35what we observe is that
• 07:36prior to treatments, regulatory t
• 07:38cells have a very distinct
• 07:40suppressive phenotype.
• 07:42And what's shown here are
• 07:44all of the regulatory T
• 07:46cell clonotypes, which are confined
• 07:47to one phenotypic cluster.
• 07:50But following treatment, we actually
• 07:51see increase in the clonal
• 07:53overlap across different
• 07:54c four effector populations.
• 07:57And this is interesting to
• 07:58us because we would hypothesize
• 07:59that disabling reg TIGIT signaling
• 08:02would destabilize regulatory t cell
• 08:04phenotypes.
• 08:06And so, this is,
• 08:08obviously in
• 08:10preliminary data and data that
• 08:12we're still working on generating,
• 08:13but,
• 08:14it does play in line
• 08:15with the preclinical,
• 08:17hypothesis
• 08:18going in,
• 08:19in a regulatory t cell
• 08:25manner.
• 08:26But what we're also interested
• 08:28in trying to understand is
• 08:29whether there are bidirectional interactions
• 08:31that are occurring. And so
• 08:33a very talented graduate student
• 08:34in our group is currently
• 08:36working on what is the
• 08:37bidirectional signaling
• 08:39that is occurring and has
• 08:41generated some very interesting data
• 08:42to suggest that regulatory T
• 08:44cell interactions
• 08:46with glioblastoma cancer cells actually
• 08:48causes shift towards more mesenchymal
• 08:50like aggressive phenotypes
• 08:53and that this
• 08:54interaction is regulatory
• 08:56Treg specific
• 08:58and can be reversed with
• 09:00t shirt blockade.
• 09:02But that environmental context matters.
• 09:04And so
• 09:05when thinking about the heterogeneity
• 09:07within
• 09:08the solid tumor microenvironments,
• 09:10we also take into consideration
• 09:12metabolic and,
• 09:15other
• 09:16other conditions such as hypoxia.
• 09:18And that potentially, this may
• 09:19lead to avenues for additional
• 09:22therapeutic targeting.
• 09:24And so with that, I'd
• 09:25just like to thank,
• 09:26the patients and families who
• 09:28have made this work possible
• 09:29and all of our, human
• 09:31work possible
• 09:32in addition to, doctor David
• 09:34Haffler, who has been my
• 09:36mentor during my time here
• 09:37at Yale.
• 09:38In addition to,
• 09:40the folks in our group,
• 09:41the greater Hafler lab group,
• 09:42and all of our collaborators
• 09:43around
• 09:44the medical school, including,
• 09:46the support of leadership within
• 09:48the cancer center.
• 09:49And,
• 09:52well represented here is also
• 09:54leadership within the core facilities
• 09:56who have also helped us
• 09:57make, this possible, including Waelin
• 09:59and Leslie.