Translational Research in Breast Cancer: Emerging Role of Immunotherapy, Novel Targets and a Hidden Dimension of Cancer Biology

June 16, 2020

Yale Cancer Center Grand Rounds | June 9, 2020

Lajos Pusztai, MD, DPhil

Information

ID: 5331
To Cite: DCA Citation Guide

Download Transcript

00:00Both in cancer and and obviously
00:03immunology and response to viruses.
00:06And we'll start with our first speaker,
00:09doctor lash boost.
00:10I allows is, you know,
00:12is a professor of Medicine at
00:14the school of Madison at the US.
00:17Will Medison Co.
00:18Leader of the genetics,
00:20genomics and epigenetics research program
00:22and director of Translational research
00:24for breast cancer in our Cancer Center.
00:26Having received his medical degree
00:29in Budapest and his subsequent
00:31doctorate at the University of
00:34Oxford blouse is really committed.
00:36His career to really understanding the
00:39biology of breast cancer and leveraging
00:42that understanding to fundamentally
00:44improving our way to deliver more efficient,
00:48more effective,
00:49and more successful clinical care.
00:51You know,
00:52through his laboratory work,
00:54his work in translation,
00:56Medison And frankly,
00:57his leadership of clinical
00:59trials in clinic research,
01:00which have not only investigated new
01:02drugs but also leverage new technologies
01:04to better define and predict how
01:07women will best respond to therapy.
01:09He really has been the triple thread
01:11and actually have academic Medison,
01:13and we're pleased to have him today
01:15to share his work in translation and
01:18research in breast cancer so loud.
01:20Thank you for Evergreen speak today.
01:23Thank you Charlie, for them the
01:25opportunity to to give a brief update
01:28on some of the translation of projects
01:30that we actually have been involved in
01:32the past few years here at Yale. Um?
01:42So this is my a disclosure slide
01:45and I would like to cover 3 areas.
01:49When is exploiting the emu micro environment
01:51of a breast cancer for therapeutic purposes,
01:54some potential metabolic vulnerabilities
01:56in breast cancer I think exists and I'd
02:00like to challenge you to think and you
02:02knew very about cancer predisposition
02:05or cancer risk category cancer risk.
02:07So the road from an idea to a clinical trial
02:11or clinical trial result is really very
02:14long and often times very sort of tortuous.
02:17So as an example,
02:19about 10 years ago we published this
02:21paper very show at the presence of
02:24immune cells in primary tumors,
02:26either ER positive or triple negative,
02:28or her two positive breast cancer service
02:30usually associated with prognosis.
02:32Let's outcome in patients receiving
02:34order surgery.
02:34Now the treatment and surgery so
02:37that people was published in 2010.
02:39I'd like to remind you that in 2010
02:41there is no effective, you know,
02:44therapy in any disease type.
02:45So we cause this the closing sentence.
02:48This paper was, well,
02:49it remains unknown whether the
02:51Indians selectivity is simply
02:52associated with with a better outcome,
02:54or it's really the cause of
02:56the battle outcome in disease.
02:58So around the same time in another
03:01project with a postdoc in my lab,
03:03we also looked at what sort of biological
03:06processes or or gene expressions,
03:08signatures or patterns are
03:09associated with the response.
03:10Show me legend.
03:11Chemotherapy.
03:12Sony rejoin chemotherapy is really a
03:14perfect setting where you can correlate
03:16a particular biomarker with sensitive
03:17to the treatment because you can
03:19directly measure the effect of the
03:21treatment at the time of the surgery.
03:23Patients have no residual
03:24cancer after the chemo.
03:25They tend to do really,
03:27really well and we call this
03:29pathological complete response.
03:30So be selling.
03:31This is again a number of immune signatures.
03:33Are Indian related markers fell
03:35out so obviously the big question
03:37is so is this a mere Association
03:39or A cause and effect and?
03:41Right around that time,
03:42in the second half of 2010,
03:45the 1st paper came out in the New
03:47England Journal of Medicine and eating
03:49woman at that have shown that actually
03:52manipulating an immune checkpoint,
03:54so we'd illumi map could improve the
03:56survival of patients in metastatic Melanoma,
03:59so that really created an opportunity to test
04:01this Kozera fact versus Amir Association.
04:042012,
04:04which is actually the year
04:06when I joined the air.
04:07We proposed the new edgmont trial to be
04:10an S and ask them would repeal me map.
04:12It could improve the pathological
04:14complete response rate when
04:15combined with chemotherapy.
04:16So PMS said, well,
04:17now that's a good idea,
04:19but it's really too toxic and
04:21it's way too early.
04:22So it went to look and proposed them
04:25the same thing and you ain't even prior,
04:27but also for good measure,
04:29he asked them whether they would
04:31be interested in Edgmont Prial,
04:32since since a lot of immune cells
04:34in the microenvironment predicted
04:35better outcome of the surgery.
04:37So for the new regiment they actually
04:39sent me this lovely letter that I
04:41thought it's interesting to read, right?
04:43So read the actual date first,
04:45so it's 18th of September 2012.
04:47The ordinary nice,
04:48polite rejection that well,
04:49we are unable to provide
04:50either funding or drug for this
04:52project because of the unclear.
04:54FDA regularly path forward.
04:56Either there is well,
04:59sort of tried the road for edge
05:01event registrations and they agreed
05:03to do an agent trial which led to
05:06this spoke as 1418 konnakol trial,
05:08so not as 40 actually means
05:11that it was started in 2014,
05:13so it took about two years to the NCI
05:15to really put together this large of
05:18the randomized registration trial.
05:20After Merck supported it in
05:222012 in principle.
05:23So I'm kind of tenacious and not
05:26very aggressive, but I I tenacious.
05:28So I supposed to say my dear again
05:30to Med immune in the same year,
05:33maybe even at that time was
05:34a startup company.
05:35They had do volume app and now they're
05:37part of easiest is Erica and being
05:40smaller than they actually said?
05:41Yeah, why not?
05:42So we propose them to studies
05:44and they agreed to both.
05:46What was the single agent,
05:47I mean a single arm phase, one phase,
05:49two trial at Yale, and the randomized trial.
05:52So the I spy consortium combined it all,
05:54Apple event added to pocket back.
05:56So our study at year was just simply.
05:58So.
05:59Hum.
05:59So I could be presented last
06:03year the results of the year
06:06phase one phase two trial.
06:09There's a historical interest because
06:10this was the first knew edgmont even
06:13oncology trial anywhere in the world.
06:15So the first patient was enrolled in 2015.
06:17It was a phase one phase two trial,
06:20and because this has never been
06:22done in the collective setting
06:23combined with chemotherapy food,
06:25though,
06:25is chemotherapy before surgery FT
06:27require that the first three patients
06:29is watch for 9 months each week
06:31for the next lot could be entered.
06:34So it looks like it was a long
06:36time to complete.
06:37The results showed that the
06:39pathological complete eradication
06:41rate in the rest was about 44%.
06:43The same chemotherapy regimen
06:44in a similar trial,
06:46and by by the sea by smog.
06:48Southwestern college group resulted
06:49in a PCR or pathological covered
06:51response with different 29%.
06:53So we also noted that there,
06:55as you would expect,
06:56you cancel each other higher PD
06:59leg and one expression or more
07:01sites at a higher PCR eight more
07:03closer to 60% rather than.
07:0544 So remember,
07:07parallel with this and the other study was
07:10running through the I spy that I was so.
07:13Lad and he showed the results of the
07:16Plenary session of the EC are this year on.
07:19The study shows 9 the randomized setting
07:21that indeed leave this door vulnerable.
07:23A product combination improved the
07:25pathological can't response rate
07:26in both the triple negative group,
07:29which was the results were
07:30very eerily similar in terms
07:32of the PCI rate, 47 versus 44% in
07:35our little negative study at Yale,
07:37and we also notice that
07:39actually this drug also work.
07:41That is combination worked.
07:42It was added to chemotherapy. You have to.
07:45I mean it hormone receptor HR stands
07:47for hormone receptor resolution,
07:49receptor positive disease.
07:50However, the result might this reason
07:53why this was picked for a plenary
07:55session is actually in the next slide,
07:58so we stumbled upon a remarkably sort of
08:01simple and clear way to isolate out among
08:04the estrogen receptor positive tumors,
08:06the ones which really benefited
08:08versus those who did not.
08:10There was no editing benefit.
08:12So if you split these estrogen
08:15receptor hormone receptor positive
08:16group into two more likely subtypes,
08:19let me call your MP1 and MP2MP one
08:22this crap sort of Mamma print.
08:25The lower end of the moment I grew
08:27up in the MP two is the Momma Prince
08:31Super High score on my printer.
08:33Similar sort of predictor that we used
08:35to to identify patients who benefit
08:37from Edgmont chemotherapy in hormone
08:39receptor positive diseases score.
08:41So what we're showing here that this
08:43core itself has a meaning and just
08:45being called high risk or or benefiting
08:47from chemotherapy is valuable,
08:50but you actually can also split this look
08:52into that really super sensitive to adding.
08:55Uh, in Indian checkpoint
08:57inhibitor to the chemotherapy.
08:58So in that group,
09:00the pathological CR equals 64% versus 22%.
09:03In the chemotherapy alone.
09:05So what is this MP group?
09:07So this NP group actually is
09:09the group which has a very high
09:11proliferation and the relatively
09:13low estrogen receptors signaling,
09:15or estrogen receptor sort of activity.
09:17Read out that you can capture
09:19Biostar generating wearing jeans,
09:20and that's an important sort of
09:22piece of information to design the
09:24follow-up registration trial for
09:26base that we have working with.
09:28Because this guy proliferation
09:30most regions signaling group is
09:32the group that is the least likely
09:34to benefit from endocrine therapy.
09:36Gotta hand it most likely to
09:38benefit chemotherapy,
09:38and we think that this benefit
09:40could be further augmented by
09:42by adding indoor follow map.
09:44So we want to get back to this
09:46letter in September 2012 from work.
09:49So a year later, in September 30th,
09:512013,
09:51the FDA approved the first sort of
09:53drug to be based on pathological
09:56company response rate in breast cancer,
09:58and that was purchased.
09:59So purchasing have improved the
10:01pathological computer response rate
10:02in her two positive disease and
10:04lettered registration of this drug.
10:06So not respecting the idea,
10:08and they actually lounge the large
10:10randomized trial with a pathological CRS,
10:12their endpoint, and to their credit,
10:14they invited me back.
10:16To the app.
10:18Leadership of the trial and the
10:21results were actually published this
10:23year in the new invention of Madison
10:26because it did confirm that indeed,
10:28adding parallelism after chemotherapy
10:31improves the pathological computer
10:33education rate improved the
10:35recurrence free survival even after
10:3818 months of median follow up.
10:40So this is the research
10:42study that sort of was
10:43largely based on these observations.
10:45Remember 10 years ago, so it took six.
10:48It looks 40 years to actually start this
10:50study another six years to complete it,
10:52and it would have been completed by now,
10:55not for the coded.
10:56So it has accrued 923 patients out of 1000
11:00and the results probably will be become
11:02available in the next two to three years.
11:05So so some cancers that high,
11:08you know, you know.
11:10Union presence know why
11:11so often suffer numbers.
11:13Medical student at Yale took on
11:15this project to actually look
11:17into the molecular background,
11:19or why sometimes have a
11:21lot of lymphocytes at this.
11:22Last one is now medical country
11:25fellow at Sloan Kettering.
11:26So we did these families pcga let
11:28me show that many other people
11:30did before that triple negative
11:32cancer's had a higher notation count,
11:35highly antigen mode and more
11:37cytotoxic T cells.
11:38However,
11:38when you look at the actual group On its own,
11:42like triple negative disease or ER
11:44positive cancer, her two positive cancer.
11:46These associations suddenly flip.
11:47This is a correlation matrix.
11:49I don't expect you to see the numbers,
11:52but the colors indicate you the
11:54the positive correlation when
11:56it's Brown and it's blue,
11:57it's negative and anti correlation
11:59and the deeper the color,
12:01the higher the correlation value.
12:02So you can see that the Indian signatures
12:05are highly correlated with one another.
12:07But on the other hand,
12:09is Genomic metrics of generic
12:11complexities such as.
12:12Mutation load or New Antigen Lord
12:14or deletions or amplifications,
12:16loads or como complexity.
12:17We actually are inversely associated
12:19with immune Indian presence in
12:21triple negative disease and disease,
12:23so that was pretty counterintuitive.
12:25Counter intuitive in 2017.
12:27OK,
12:27so you're moving in says you have a simpler.
12:31We can suggest the fewer than
12:33your hand surgeons are.
12:34So when I see things like this,
12:37I try to confirm it.
12:39So we reached out collaborator and
12:41friend Thomas card and ask him to
12:44actually replicate this or with a
12:46different methodology and Thomas
12:48actually find exact same thing that
12:50in primary triple negative disease,
12:52the fewer the immune cells they hire.
12:55The Genomic heterogeneity and
12:56the worst prognosis. So worse.
12:58Prognosis means that you have a
13:00higher probability for methods.
13:02This is so and let us do this hypothesis
13:05that maybe the metastatic lesions
13:07are actually immune refugees or or escapes.
13:10So there are more immune in Earth
13:13that Michael Environment is more
13:15immune inert and these three
13:16brilliant woman took on this project
13:19or parasitically was a visiting
13:21scientist from Hungary or so with.
13:23Smoking baby that his lab in Charlotte.
13:27With me and.
13:28More Gerstein so we rounded up samples
13:30which will pair primaries and meds and
13:33also a separate Court of primaries.
13:36And that's not from the same patient,
13:38and subjected them to a whole series
13:41of molecular studies to test it in.
13:43You micro environment is the
13:45same or different,
13:46so this is just really
13:47examples of three sort of
13:49simple and straightforward
13:50findings with teal comes the tumor.
13:52Infiltrating info sites are lower in
13:54maximum primary tumors in matched
13:56and unmatched chords be dealing.
13:58Expression is the same, it slower.
14:00In in the matter equations and
14:02also a whole lot of different
14:04Indian signatures are all lower,
14:06consistently showing that there are less
14:08Indians and less activity in the breast.
14:11Cancer metastatic micro environment.
14:12Why don't was really interesting though?
14:15Is that while most of the
14:17emu markers went down,
14:19some of the Indian targets actually
14:21remain high or or even increased
14:23in the meta static environments.
14:25And these IO targets are potential
14:27good set of candidates for offer.
14:30Testing them in the meta static setting,
14:32either alone or in combination.
14:34Also in combination with established agents.
14:37So we selected the group of this
14:39preserved IO targets for a clinical trial
14:42that we hope to conduct this support.
14:45Clinical trials it's called rustic
14:46and this is sort of a scheme of it.
14:49And again,
14:50this takes this immune targets that we
14:52preserved in the meta static setting
14:54and testing in the clinic whether
14:56they really have a functional role.
14:59In suppressing the new Micro Hood,
15:00so because of lack of time I
15:02can't really talk about all the
15:04other Indian projects every day,
15:06but I just thought I list them here
15:08so we did compare changes in the
15:10micro environment before and after
15:12therapy and the shouts only published.
15:14This give us some ideas what to add
15:16to Pembrolizumab Order Volume app to
15:18make the treatment even more effective
15:20than you actually and setting.
15:22Homes we also compared to same
15:24day the immune reach,
15:25triple negative and the energy or
15:28positive Kansas to see that there is
15:30differences in their micro environments
15:32and that was done by Paso mirror
15:35medical students who is now a resident
15:37at the Harvard system the same way.
15:40So we did similar comparisons
15:42by by race and King blindness.
15:44Scientists in my lab is working or not.
15:47Of data from a number of different new
15:50agent trials that seems to do well.
15:52Map kind of what precisely
15:54defined predictors,
15:54and he rosenblit is Medical College
15:56of fellow with a very nice people and
15:59actually looking at in a large pool
16:01of data from Foundation Medicine.
16:03Be like an expression across different
16:05meta static sites in breast cancer.
16:07And there are some really substantial
16:09differences in people like an
16:11expression depending on what site
16:13you are actually sampling.
16:14I'm going to move on to something else
16:17that really got me excited in the past.
16:19So if you a few months so.
16:21Many metabolic processes are catalyzed
16:23by multiple different isozymes or or
16:25proteins that really capitalizes.
16:26Same enzymatic reaction.
16:27Normal cells usually have many of these,
16:30and oftentimes in cancer you actually
16:31see that one of the isoforms become dominant,
16:34so that's the schema on this figure, right?
16:37So normal tissue is both sides.
16:39When I sent to our expressed in cancer,
16:41I just have one becomes a dominant
16:44and the other
16:45one is lost. So we asked how often do we
16:47see this in cancer and do this sort of isis
16:51and expression changes could could harbor.
16:53Or or include enzymes that we could target
16:57metabolically somaca March secret visiting
16:59scientist faculty from from a Polish
17:02University to this project on and device.
17:05This sort of strategy to
17:07look at humanizing forms.
17:09Isozymes managed matched primary
17:11tumors in the metastatic lesions
17:14sorry match the normal tissue.
17:16The DC area across 14 different cancer
17:19types where this was available than the
17:21validated the results in cell lines.
17:23Make sure that this is really
17:25observed in the purest system and
17:27not just an artifact of difference.
17:29He still different issue compositions
17:31and then says the functionality
17:33in this depth map data which is
17:36basically complete knockdown of all
17:37human jeans in about 7 or 8 cell
17:39lines and then the conference hits.
17:41We validated in the manual screen.
17:43This is an example for you
17:45how this exactly looks.
17:47404 Kansas so this is Csea enzyme.
17:522 forms ACA&B.
17:53Plots show you that how they still the ACA,
17:57which is the red and be which is the blue,
18:00had the expression a normal and an action.
18:03Cancer tissues and you can see that
18:05the Blues all go down and cancel.
18:07That means that the expression
18:09of this is lost,
18:10whereas the red remains stable
18:11and the red is the sea.
18:13So this thing this is a
18:15potentially interesting target.
18:16So when you look at this across
18:18different cancer types and
18:19indeed app map validation data,
18:21then actually what really fell out on the
18:23top is this questionnaire carboxylase.
18:25Which show this loss of isoenzyme
18:28diversity intro different cancer
18:30types and was socially but nine
18:33different cell lines and each cancer
18:35types in in the depth map so that
18:38map has like several dozens of cell
18:40lines for a particular cancer type,
18:43like breast cancer.
18:44Nine of these showed the same
18:47loss of heterozygosity,
18:48loss of diversity as we
18:50saw in the human data.
18:52Different cancer types.
18:54And in this case is also validated,
18:57so most of the cell lines there
18:59where the CSC was knocked out.
19:01If it had the dominant expression,
19:03it really impact viability.
19:04But the real kicker is that when we
19:07look up what is known about this,
19:09it turns out that Pfizer has a drug that
19:12they put through phase one and phase
19:14two trials for diabetes and fatty liver,
19:17and actually showed all the
19:19metabolic effects that we expected.
19:20But they discontinued development
19:21last year or two years ago
19:23because of thrombocytopenia,
19:24which is wonderful.
19:26Because don't beside opinion
19:27through megakaryocytes really rely
19:29on the normal lipid synthesis,
19:30because from both sides bought off and
19:33every time I turn both sides come off
19:36from the surface supermodel career site,
19:38it takes lipids.
19:39Membranes made it so we answered
19:41the proof that is really works
19:43the way it's supposed.
19:45So the Anthony to collaboration with
19:47Pfizer to to do some additional
19:48preclinical studies and bring
19:50it in the clinic
19:51if it validates so,
19:52we simply throw validation is pre
19:54clinical validation is falling on
19:56the shoulders of Julia fold even
19:58offer Medical College of Fellows and
20:00finish with scientists in my lab.
20:02So before the coveted broke we had a
20:04chance to look at 10 different cell
20:06lines and you see that in the human
20:09sort of achievable concentrations
20:10that you can get in the human serum.
20:13It says a pretty broad inhibitory effect.
20:15And the army is not for coffee.
20:18I could probably show you,
20:20said the combinatorial screen results
20:22from the high throughputs combinatorial
20:24screen that we initially is doing in
20:27our core facility at the West campus
20:29and also collaborating in jacks
20:31to test this drug in PDX models.
20:33And we hope to bring this to the clinic.
20:37Maybe the year 2.
20:38So finally the last five minutes I
20:41wanna spend on an idea that we kind
20:43of stumbled upon off awhile back.
20:45This is not our paper,
20:47it's the people from nature of it shows
20:49you the distribution of different sort
20:51of mutations in large cities of Kansas.
20:54So the striking thing about this
20:56is that there are these set of
20:58jeans that affected more than 65
21:00cases out of close to 3000 Kansas.
21:02And even in this sort of very modern
21:05and high sort of accuracy study,
21:07about 9% of Kansas said
21:08no driver alterations.
21:09English challenge you to think about anyway.
21:11What you think you mean by a driver, Jean.
21:14So is it a statistical construct
21:16from any sort of statisticians
21:17in computational biologist?
21:19It is actually a statistical construct,
21:20but of course many of you think about
21:23this is gene that caused the cancer.
21:25The way I think about this is
21:28actually it's just a narrative tool.
21:30To kill a good story.
21:32So this is actually from the same paper
21:34from but from the supplementary figures,
21:37but it shows you the enormous amount
21:39of model of Genomic abnormalities
21:41that any particular cancer
21:42has so retro transpositions,
21:44a few dozen number of structural variants,
21:47several dozen to several, several 1000.
21:49So these are big chunks of the DNA
21:51chromosomes missing very larger than
21:53the thousands of in Dallas and 10s of
21:56thousands of single included variance.
21:58Incidentally,
21:58you also see this actually in the CIS,
22:01which is a premalignant lesion,
22:03so these services, all the.
22:05Well,
22:05marks off of cancer,
22:07except it's not really cancer,
22:08but it has the same B 53 mutations
22:12clarifications or not.
22:13Just the game keeps you big pools,
22:15better weather really.
22:16The function of this these jeans
22:18and then the individual jeans is.
22:20So this is a people that many
22:22years ago we did nearby Sunday.
22:24She was that he was a medical student.
22:26That year.
22:27Now is a faculty at Sloan Kettering
22:29and what I want to illustrate here
22:31is that every single cancer which
22:32is a column as really a different
22:35combination of abnormalities.
22:36So if you think about it that way,
22:38maybe it's really the reason.
22:39Why cancel the have different layers?
22:41Because because of this combinatorial
22:43difference that they have.
22:44So if each of these contributes
22:46something then their net effect is
22:48really really heterogeneous behavior.
22:50But maybe it's even more interesting.
22:52Is this work with DVR? She was.
22:55The students at that time at Yale
22:57and now it's a medical student pad.
23:00You sequence all the human kindness
23:01ease in 90 two breast cancer,
23:04only to see whether there are any.
23:06Lowering additional kindness is that it?
23:08I guess we didn't find any,
23:10but we really observe though
23:11is that there is a very large
23:13number of high functional impact.
23:15Variance in kindness is bigger germline.
23:18I'm still think about for a second,
23:20so you actually carry a bunch of germ line
23:22so the mutations that inactivator overactive.
23:25It kinda seems like PSC kinase or whatnot.
23:27So what does it mean?
23:30So please give us this idea
23:32that maybe it's really.
23:34He just focused too narrowly
23:36on driver mutations,
23:37which are only four of five in a cancel.
23:40What actually would be probably also
23:42helpful is to look at the context
23:44in which this is happening and
23:46the constant Israeli hundreds of
23:48additional variants that come in from
23:50the somatic or the germ line angle.
23:53So he proposed this idea that
23:55really functional German variance,
23:57conkle, jeans and it's the totality
23:59of the functional impact.
24:00High functional impact German
24:02variants in cancer Lady Jeans.
24:04They could actually determine cancer risk.
24:06So we know that there are
24:08a few very high penetrance.
24:10Cancer is chains like Bronco,
24:12but it's really is the minority
24:14women who carry this even very,
24:16very strong family history.
24:17So what's accounts for dismissing heredity?
24:20You think it's the totality of
24:22the defects that actually are.
24:24Embedded in a whole lot of
24:26individually non sort of.
24:28Cannot translate that.
24:30And then the next iteration of this,
24:33that's really the combined effect
24:35of the germline and somatic events
24:37that really lead to the malignant
24:39transformation rather than a few
24:41individually dramatic effect.
24:42And it's a project that talking
24:45is pursuing in my lap towers
24:47visiting post doc from from.
24:49And that is so if it's really true,
24:52then we would expect that woman who
24:54developed cancer the younger age will
24:56have a lot more sort of deleterious
24:59germline events in cancer jeans then
25:01people who develop cancer the later
25:03age because it's ultimately the
25:05combined effect of the acquired and
25:07the inborn errors that lead to Kansas.
25:10So if you are born with a
25:12lot of errors to start with,
25:14it gonna take a fewer or shorter time
25:17to get to reach this critical level.
25:20And long, behold,
25:21that's exactly what he observed in a
25:23bunch of large series like that ECA or
25:25the UK biobank and now just published
25:28this paper in Nature Communications.
25:30I want to just point out to you because
25:33time is short on this very last figure,
25:36which is the the comparison,
25:38the relationship between the
25:39mutation than in the cancer versus
25:41the variant, the German
25:43variant burden by age groups.
25:45So he he means younger than 30 and the
25:49other advocating speak louder than 80.
25:53The installation ship is actually remarkable,
25:55so yo patience young who are who
25:58have cancer at a younger age and
26:00it's it across all the cancers that
26:03letter TCG ahead or the UK biobank.
26:05We also get consider the 30s, forties,
26:0850s have a much higher germline
26:10variant burden in cancer Lee jeans,
26:12and this is like 5 or 6 only jeans then.
26:16Then people who get cancer the older
26:18age and on the other hand most
26:21folks have a much higher mutation
26:23somatic mutation so they can.
26:25So so then that led to us
26:27another idea that So what?
26:29Actually the cancer jeans are in this thing?
26:32That it's probably a lot broader
26:34sort of repertoire then then
26:35no one Canonical cancel jeans.
26:37So you could think that this study?
26:40The Dom Hussein,
26:41who is a PhD student in the
26:42computational biology program and
26:44supervised by Mian Mar Gerstein too.
26:46So how many jeans are actually connected,
26:49one step or two step or three step
26:51away in a pretty important interaction
26:53network from from a cancer gene.
26:55So if they are immediately next to it,
26:58then they probably influence
26:59the affective for cancer gene.
27:01If they're two step removed,
27:02they still probably influence over less,
27:04so there actually a whole lot of jeans, but.
27:07Half of all human jeans are
27:09actually connected.
27:10One or two steps away from
27:12from cancer hub gene.
27:13Most of these are not implicated in cancer,
27:16and if you look at their sort of
27:18functional importance in this gap
27:20database and it turns out that the
27:22further away you get from the cancer hub,
27:25the less important they seem to be
27:27in survival in the depth map data
27:29which supports the idea that there
27:31are lot more jeans involved in
27:33cancer than what you think you see
27:35the same when you look at weather.
27:38There is an evolutionary pressure to to
27:40preserve truncating mutations in these jeans,
27:42and again the further away.
27:43So the jeans which are one Step
27:452 step three step away from a
27:47concert Hall gene and less and less
27:50evolutionary pressure on them into
27:52four through 4 to exclude truncations.
27:54And it was so carry actually
27:56somatic mutations. So that's.
28:02Personalized sort of Jim Langley scored.
28:04That's sums up all these effects that
28:06people are born with cancer chains,
28:08and that's a project that
28:10that we actually got to ask.
28:12Or young investigator award
28:13to pursue together with cow.
28:15And we say this too
28:16gruesome pictures out here.
28:18Just remind you guys that
28:19this is how airplanes crash.
28:21They actually don't crash because
28:23there is a statistically significant
28:25losses of the wings or the engines
28:26or not every single plane crashes
28:28caused by a different combination
28:30of individually nonlethal events.
28:32So they fall into groups like human error,
28:34which is almost always there,
28:35but it's not the same human error.
28:37It's a different kind of human error.
28:38There's always some kind of a
28:40mechanical or instrumental failure,
28:41but it's never the same instrument.
28:43So that's my final slide,
28:45really studying the new micro environment.
28:47Let us to some some useful ideas
28:49about clinical trials are very
28:51excited about exploring metabolic
28:52adaptations for therapeutic targets,
28:54and we submitted the DOD grant with this
28:56with a group of other investigators,
28:59and I think I really think that the
29:01universe is functionally cancel.
29:03Event chains is much larger than
29:05we think it is interested in.
29:07More stuff at the bottom of the
29:09slide shows takes you through the
29:12list of publications by our group.
29:14So thank you, and this is my lab.
29:18Each other.
29:21Wow, thank you.
29:22That was a an impressive array
29:24of work on so many fronts.
29:27And congratulations on all of it.
29:29I know where we're a little late on time,
29:32so let me just offer up one question.
29:35You know? I think you oppose.
29:37Obviously a very good case that
29:39it's a combination of germline
29:41asmatic events and I'm curious,
29:43do you think breast is is different?
29:46Breast cancer is different than other
29:48solid to malignancy's because obviously
29:50germline and semantic events are.
29:52Install cancers,
29:52but you think breast is
29:54somehow different that respect.
29:55Yes, it's different. It's a matter of fact,
29:58so you can group really Kansas even
30:00in this paper that that I refer to
30:03be looked at different cancer types,
30:05and this Association starts to fall apart
30:07in Kansas that actually have a very high
30:09environments or customer on exposure, right?
30:11Because that in this sort of
30:13message of this new relationship.
30:15So this relationship is less strong
30:17in lung cancer, bladder cancer,
30:19and some other cancers.
30:20So the real picture of course, is nuance.
30:23It's much more nuanced.
30:24And the same way.
30:25So the jeans we chat important so the
30:28cancer gene is probably also vary from
30:31some tissue types of tissue type.
30:33So these are the refinements
30:34that we are actually working on.
30:36Is just that I wanted to give you
30:38a repertoire of things that we do,
30:40but that's exactly what we actually
30:42addressing in this project right now.
30:44Thank you and I know just for time
30:46will will move on, but obviously
30:48folks can certainly email allow us to.