"Thoughts on the Nature of Breast Neoplasia"

January 25, 2022

Information

Julia & Patricia Kingsbury Memorial Lectureship/Yale Cancer Center Grand Rounds | January 25, 2022

Presentation by: Dr. Larry Norton

ID7376

To CiteDCA Citation Guide

00:00You can see it. Fred.
00:04So with no further ado on
00:06behalf of Doctor Nita,
00:08Hoosier Interim Cancer Center director,
00:10and actually this time next week,
00:12your good friend Doctor Eric
00:14Wyner will be sitting in that
00:16position as our permanent director.
00:18We're really excited to have you here,
00:21even though it's not in person,
00:23as the Julia Patricia Kingsbury
00:25Memorial lecturer and lectureship
00:27that's been sponsored for.
00:28You know well over 2 decades by
00:31their family for me to introduce.
00:33Doctor Norton is like.
00:34Introducing a Rockstar.
00:35Obviously he's a senior vice president
00:37in the office of the President,
00:40Memorial Sloan Kettering,
00:41the medical director of the
00:43Evelyn Lauder Breast Center.
00:45I think you're also the founding
00:48and incumbent N Norna Serafin
00:50clinical chair in oncology.
00:52Career started,
00:53you know,
00:54undergraduate and undergraduate at Rochester,
00:56then went on to Columbia for medical
00:58school in Albert Einstein and the NCI
01:01for Training and Medicine and Oncology.
01:03And really,
01:04you know the first time I was I was
01:06privileged to meet Doctor Norton
01:08was in 2002 at the old CLG or
01:11the cancer and Leukemia Group B.
01:12And you know that committee which
01:14you chaired for such a long time and
01:17then passed on to doctor Doctor Weiner again.
01:20Our incoming director and Doctor
01:22Hudis just to see how masterfully.
01:24The research,
01:25the work,
01:25the care of your patients over the years,
01:28and then you know more recently in the
01:30last decade you know being involved with
01:32the breast Cancer Research Foundation,
01:34which you and Evelyn Lauder,
01:37the late Evelyn Lauder,
01:38and Leonard Lauder,
01:39you know,
01:40put together really bringing over 200,
01:43I think,
01:43probably close to 300 of the
01:45world's leading investigators
01:46and breast Cancer Research.
01:48Really, for the cure,
01:50as defined as the founding
01:52scientific director.
01:53Gosh,
01:53this is such a.
01:54A great day for for a Yale and I
01:57know everyone is really excited
01:59to hear your thoughts on the
02:01nature of breast neoplasia.
02:02So thank you Doctor Norton
02:03for making the time.
02:04OK, thank you. Thank you very much.
02:06I hope everybody can hear me and thank you
02:08for that really very gracious introduction.
02:10You know it's a it's it's totally shame.
02:13In the old days when you give a lecture
02:14ship like this, you'd come in person.
02:16You'd have a dinner you'd meet with a
02:17lot of people, one on one, and so many
02:19of my great interactions in my career
02:21started really by those kinds of events.
02:23And so it's a. It's a shame that
02:25we have to do this electronically.
02:26But it's a great pleasure to be
02:28here and and speak with you.
02:30I'm, you know,
02:31my neighbors in the Northeast about some of
02:33the things that that I've been thinking of.
02:35What I've been doing,
02:37it's called mathematical insights,
02:38but for those of you who are math phobic,
02:40please don't don't run away screaming.
02:43You know it's a.
02:44I'm only going to show one equation,
02:47and it's not important really for the talk.
02:48Basically, it is mathematical thinking
02:50and a lot of people don't know math.
02:52Don't realize that what math
02:54is is is not the equations.
02:56The equivalency would be sheet
02:58music for music.
02:59The sheet music is not the music,
03:00it's the sound.
03:02And and with mathematics it's it's
03:04the the insights that you gain which
03:05you know in terms of how things.
03:07In this case, how they grow,
03:08how they shrink,
03:09why they grow that way and and and so on,
03:11how we take advantage of that.
03:12The the equations are not really the
03:15mathematics for many years when I was
03:17giving this talk I I skipped over a
03:19lot of the early stuff that I did,
03:21but then I realized a few years back
03:23that a lot of the younger people
03:25are unaware of that early stuff.
03:26And that really stuff is really
03:28very important for understanding
03:29the later things that we're doing.
03:30So I am going to be talking about it.
03:33I mean,
03:33it really happened to me a bunch of
03:35years already that I was a visiting
03:37professor and somebody presented a
03:39case and and said this patient dose.
03:41Dense chemotherapy with AC and Taxol, Dr.
03:45Ordinary familiar with that regimen.
03:47And that's when I realized that
03:49that perhaps I should really cover
03:51some of the early things that I
03:52that that I've done and and how it
03:54relates to to the bigger picture.
03:56So I'm going to talk.
03:57About growth models and of
03:59course the the premier growth
04:00model being from Howard skipper,
04:02I'll talk about the work that I did
04:05in the 70s interpreting that growth
04:08model in with the appreciation for
04:11understanding a different pattern of
04:12the way that cancers grow than how it
04:15skipper and and colleagues had shown.
04:17How it led to the concept of dose
04:19dense sequential therapy and what
04:20are the results of that that we
04:22just fairly recently summarized
04:24by the Oxford overview?
04:26And then talk about self seating
04:28theory and how it relates to all of
04:30that previous work and that will
04:32bring me into the area of fractal geometry,
04:34which is where where another topic
04:37in math comes in and how our that's
04:41informing our current work on the tumor,
04:45infiltrating leukocytes,
04:46and the interpretation of their firm.
04:48And I don't know if David Rim is,
04:50you know, here among us today, but,
04:52but we've had a number of early
04:54conversations a few years back about.
04:56The importance of fractal geometry
04:58and understanding biology from
04:59a pathology point of view,
05:01and then how that relates to concepts
05:04of drug resistance and and the use of
05:08immunotherapeutic agents and and lately.
05:10Our work that we're doing on antibody
05:12drug conjugates in that regard,
05:13but all informed by mathematical thinking.
05:16Let's just start back with Hippocrates,
05:18the father of us all.
05:20The parent of us all in, in in,
05:23in, in medicine.
05:24The actual quote translated from
05:25the Greek is an illness is once
05:27you keep two things in mind to be
05:29useful rather than cause no harm.
05:31That's frequently misquoted,
05:32as as first of all, do no harm.
05:34That's not quite what he said.
05:36What he said is don't be neutral,
05:37but but, but, but be useful,
05:40and that is,
05:42is is a very important quote
05:45because it relates very.
05:46Very directly to one of our major topics
05:48that we have to deal with in clinical
05:50oncology all the time, which is OK.
05:52I have a drug that works.
05:53How should I use it and dose level of
05:55course is a mix between the efficacy
05:56of the drug that you're giving and the
05:59toxicity that you're causing from it.
06:00And I spent almost all of my youth
06:02learning to be a medical oncologist
06:04learning how to to avoid or manage
06:06toxicity of of the agents and pick
06:08out the right dose level quotes
06:09in the modern world has gotten
06:11much more complicated than that.
06:13We have to not only look at
06:14at the at the dose level,
06:15but also the schedule.
06:17The duration of Therapy will give it
06:20impulses and that leads to various
06:22changes in efficacy and toxicity.
06:24Toxicity is not just acute toxicity,
06:26but late toxicities chronic
06:28toxicities that may arise.
06:30The personal cost to the patient and the
06:32personal goals for the patient have been
06:34taken into account and and in planning,
06:36dosing and scheduling as but also
06:39societal cost that everything that
06:41we do is going to have implications
06:43basically to all of our society.
06:44All of our patients and society in general.
06:47And how does all of this relate to a
06:49very rapidly evolving therapeutical
06:50landscape so so dosing is scheduling
06:53is actually a very germane topic
06:54in in the modern era,
06:56even when we're talking about
06:57some of the newer agents.
06:58And and what we've learned in
06:59looking at the older agents, IE.
07:01Chemotherapy is directly related to how
07:03we're going to be applying our newer agents,
07:06and as I close the talk,
07:07I hope I'll be addressing some
07:09of these points.
07:10But the central dogma that led most of us in
07:13our careers in medical oncology is this.
07:16If you want to kill more cancer cells,
07:18you have to use higher dose levels.
07:19So you want to use the highest possible dose
07:22level you can to kill more cancer cells.
07:24Because the more cancer
07:25more cancer cells you kill,
07:26the more benefit to the patient
07:28either eradicating the cancer.
07:29If if if that should actually be possible,
07:32or just buying time 'cause we have smaller.
07:34Buying more tumors can take longer to regrow,
07:36and that's going to be translated into
07:38into improvement in duration of Disease
07:41Control for the patient and hence our
07:43training was all about determining and
07:46treating at maximum tolerated dose.
07:50I'm going back to another great
07:53teacher of medicine, William Ostler.
07:55The greater their ignorance,
07:56the greater the dogmatism,
07:57and I believe that this this this.
07:59This dogmatism is really dominating
08:01us even to the present day
08:02when we have targeted agents,
08:04and yet we're still trying to
08:05achieve maximum tolerated dose.
08:07Thinking that we're going to be
08:08benefiting the patient by doing so,
08:09and I'd like to really question
08:11that this this concept really came
08:13from the work of Howard skipper,
08:15Franckesche Bold and Griswald and others
08:17at at at Southern Research Institute.
08:19It was an extremely important part of my
08:22education in in in medicine and oncology,
08:24in in in the 60s.
08:26The concept was based on mooring,
08:28leukemia and Howard skipper made
08:30the observation that if you if you
08:33inject a certain number of cells and
08:35the mouse died at a certain time,
08:37that could tell you basically
08:39how many cells you injected.
08:41'cause it took a certain very
08:42reproducible amount of time for those
08:43cells to reach a lethal number.
08:45So all of his work was not really
08:47measuring cancer cell numbers,
08:48it was actually measuring animal.
08:49Death and extrapolating back
08:51to cancer cell members.
08:52And that's common,
08:53not commonly appreciated is is
08:55that it was all extrapolation,
08:57but the the fundamental observation that
08:59he made is that if you kill cancer cells,
09:01you can extend lifespan and the
09:04extension of lifespan was a.
09:06Basically it took time for lethal
09:07number of cells to arrive and you can
09:09go back and extrapolate from that in
09:11terms of how many cells you killed
09:13because it would take that certain
09:14number of cells that were left or
09:16residual to lead to eventual or lethal
09:18number and and the death of the mouse.
09:21This led to a concept that is shown
09:24here by one of the one of the.
09:27Very very often in in my youth,
09:30especially reproduce figures is
09:31that if you start off with a large
09:33number of cancer cells and you
09:35give a certain dose of therapy,
09:37you kill and will go back over this
09:39a constant fraction of the cells
09:40that are present with each dose.
09:42Each chemotherapy core skills,
09:43in this case,
09:442 logs of gotta get rid of this pop
09:47up two logs of kill means means you're
09:50you're killing 99% of the cells,
09:5290% is 1 log killed,
09:5499% is a two log kill and you
09:56could drive to cure unless you get
09:58the emergence of drug resistance.
09:59Of course,
10:00if you stop treating when the cancer
10:02is disappeared, that's not enough,
10:04because because there's plenty of
10:05cancers left and they can grow
10:07back and and the concept of roses
10:09that if you start a small of tumor
10:11size that you can actually get rid
10:13of the cancer cells before this
10:14emergence of drug resistance arises
10:16and hence the concept of Azure and
10:19chemotherapy came from this Vince
10:20Davida long associated with Yale
10:22was my great teacher and still
10:24is my great teacher and
10:25in oncology took these concepts and use
10:27them to develop the MOP chemotherapy
10:29regimen and those of you who have not.
10:31Read Vince on his book of the Death of
10:34Cancer about the early days where this
10:37figure was extrapolated into the cure
10:39of a solid tumor Hodgkin's disease.
10:41You really should read it because it's
10:43it's an excellent book and it really
10:45captures the excitement of those early
10:46days in oncology and the application
10:48of this mathematical model to the
10:50development of a curative regimen.
10:52Well, this was led in the in in in the
10:5560s to the concept of dose escalation.
10:57This way if you have no therapy and you
11:00have simple exponential growth like this,
11:02you give one drug.
11:03You get certain log kill two drugs.
11:05You get you double that.
11:06If this causes one log kill and this
11:08causes one log kill then you get 2
11:10log kill 90% sale killing here and 90%
11:12sale killing here since 99% sale of
11:15killing three drugs should be should
11:17should cause disease eradication.
11:19Therefore,
11:19four drugs should certainly cause
11:21disease eradication, which was very.
11:22Influential,
11:23they're thinking about Rob chemotherapy,
11:25but it also applies to doses.
11:271 dose, 2 doses, 3 doses,
11:30all increased cell killin,
11:32causing DC radication.
11:33So the 70s was a decade of enthusiasm.
11:36Fueled by this confidence in the skipper.
11:38In the skippers model,
11:39there are many drugs that came along,
11:41such as the Cyclones,
11:43the platinum agents,
11:44the concept of combination chemotherapy
11:45as I've just demonstrated to you,
11:47arose from these thinking.
11:49And indeed we're getting successes.
11:51Cure simply nysm.
11:52And and leukemia is
11:54infamous testicular cancer,
11:55it really looked like we were
11:56moving in the right direction.
11:58Getting high response rates and many
12:00other tumors including breast cancer.
12:01The field that I eventually specialized
12:03in the concept of postoperative attribute
12:05chemotherapy rose from that period.
12:07Based on that on that mathematical idea
12:10and an enthusiasm for those those level
12:12escalation which we led to a lot of
12:15enthusiasm for a mega dose escalation,
12:18which is bone marrow transplantation.
12:20This enthusiasm was so was so.
12:23Pronounced that a mentor, not Vince,
12:26is another mentor in 1976 cents Me Larry,
12:29you still got a chance.
12:30You're young enough to change
12:31your career path,
12:32you're not.
12:32There's not gonna be any field
12:33of oncology in a few years.
12:35All these combinations.
12:35All these agents are just going
12:37to come together and cancer will
12:38be disappeared in a few years.
12:39You better think about training
12:40and something else.
12:43Well, I persisted against that
12:45advice and kept working on cancer.
12:47And as you know it hasn't been that easy
12:49and a lot of that enthusiasm is still
12:51there and we definitely making progress.
12:53No question about it,
12:54but the rate of progress really
12:56has has slowed even with the
12:58addition of of newer agents.
12:59And we're not getting cures of
13:01metastatic colon cancers and
13:02and and and stomach cancers.
13:04And and and and and many
13:06lung cancers and so on.
13:07But certainly breast cancer as
13:09readily as we would have hoped.
13:10Metastatic disease is still a big problem.
13:13So what went wrong?
13:14And this is my number one favorite
13:16quote and kind of a kind of you know,
13:19model for my life.
13:19It's not what you don't know
13:20that gets you in trouble.
13:21It's what you know that for sure
13:24that turns out not to be true.
13:26And the thing that we knew for sure
13:28was that the skipper model worked
13:30because cancers grow exponentially,
13:31but they don't grow exponentially,
13:33nor do they grow in a strictly
13:35linear fashion.
13:36And we know this because
13:37if that were the case,
13:37from the time of initial diagnosis to
13:40the time that that the cancer would
13:42cause problem would be too long.
13:44Exponential growth also doesn't
13:45make sense because for the
13:46time of initial diagnosis,
13:48the time of lethality would be too short.
13:50It's got to be somewhere
13:52in between and indeed,
13:54Benjamin Gompertz in 1825.
13:56Then invented a curve of human mortality,
13:59which we call the Gompertz curve is
14:01and kind of sigmoid curve sigmoid.
14:03Because you see as as as shape and
14:05and others had shown that that
14:07Gompertz curves actually applied to
14:08the growth of experimental tumors.
14:10I got into this in the mid 70s and
14:13this early paper I wrote in nature.
14:16In 1976 these are two rat tumors.
14:19This is a mouse tumor and what we
14:21found are working with Richard Simon
14:23is that if you have a few early
14:25measurements that it actually fits
14:27a pattern and you could predict
14:28later measurements that gone.
14:29Protein growth was really very predictable.
14:31This is really an important
14:33observation that just sort of sat
14:34there and up until the present day,
14:36but it actually is rather
14:37meaningful but nest.
14:38But but indeed,
14:39Gompertz equations are applied
14:40and then not exponential,
14:42and so my my work was basically
14:44to see how do we apply the
14:47skipper Schaible principles.
14:48To the.
14:50How do we apply skipper Schaible
14:52principles to gum persien curves
14:54and papers in the early days about
14:57this and they eventually led to
14:59the concept of sequential therapy
15:00and then dose dense therapy?
15:02And this was the work in the 70s
15:05and and again Vince DeVito was
15:07was working closely with Johnny
15:09Bonadona in those days and and
15:11some of these ideas got translated
15:13and indeed this was actually a
15:16competition in the adjutant setting
15:18between a other modelers.
15:20Called Goldie and Coldman and
15:22and myself and and Richard
15:23Simon that they predicted that if you
15:25have agents like doxorubicin and see
15:27at math you should use them in an
15:29alternating fashion because that would
15:31get this drug in these all the drugs in
15:34sooner to limit the emergence of drug
15:36resistance by random mutation whereas my
15:39modeling which I'll show you in a second,
15:40suggested that it would be
15:42better to use them sequentially.
15:43So we'll go over that modeling because
15:46this is buried in ancient literature
15:48and was published before many of you
15:50who are listening to this lecture were.
15:52Warren, so you're not aware of
15:54this work is that, of course,
15:55it's always better if you've got
15:57two agents or two combinations,
15:59you gonna use them simultaneously if you can,
16:01that's going to give you maximum cell kill,
16:03but you can't really do this in most
16:05situations without such toxicity that you
16:07have to reduce the dose levels of the drugs.
16:10And by reducing the drugs,
16:11dose levels of the drugs,
16:12you're not going to get the maximum
16:13efficacy from any of the drugs.
16:14So the question is,
16:15what can you do if you can't give
16:17them in a simultaneous combination?
16:19The Goldie Coleman idea
16:21is you alternate them.
16:22And by the Norton Simon modeling
16:24when we did it, we found well,
16:26we got a very inferior cell
16:28killed than if we did them.
16:29Obviously by simultaneous therapy.
16:31But if you give them in
16:33an alternating fashion.
16:35You can actually get cell killed.
16:37That's better than you can get by
16:38giving him an alternating fashion,
16:40so the bonadona experiment which
16:42started in the mid 80s was a was
16:45basically a competition between
16:46this approach and this approach.
16:48And of course this approach one.
16:50Then, with the advent of grants,
16:52iconic stimulating factor where
16:53you can actually squeeze the
16:55doses of drugs closer together,
16:57you can actually get maximum self
16:59kill and even a better cell kill.
17:01Then you can just with the simultaneous
17:03combination by by the application of GCSF.
17:06So you can make a through cycle,
17:08for example into a two week cycle.
17:10Tax oil can be given even
17:12without GCSF in a one week cycle,
17:14and that's also a dose dental
17:16regimen as well.
17:17So that's the understanding of dose density,
17:19which is kind of lost in
17:20history a little bit.
17:21And a lot of people don't appreciate
17:23really where it came from.
17:24Water million means.
17:26Hence we designed this regimen
17:28which basically started.
17:30In 1997, nineteen 9741 with Mark Citron,
17:36who we just lost very recently
17:38from a from from a neoplasm.
17:40A great great Lawson,
17:41and really a great clinician and
17:44and and a great clinical scientist.
17:47And this was A and the regimen
17:49was a two by two design,
17:51and you'll notice those of you who
17:52are doing cooperative group studies.
17:53We don't do two by two designs
17:55very much anymore,
17:56and I really wish we did because
17:57it would answer a whole lot of
17:59questions faster than doing doing just
18:01some of the some of the
18:02regimens that we're now doing,
18:03which is comparing 2 treatments or
18:05now even not comparing it at all, but
18:07basically comparing it to historical data.
18:09Which is, you know, Noninferiority designs.
18:15For the topic about the the wisdom of that,
18:17but, but certainly certainly these two
18:18by two designs get a lot of information,
18:21and we gave a. We gave adriamycin and
18:25cyclophosphamide AC with paclitaxel.
18:27Either all the drugs in in in in a Q3
18:32week regimen, in a sequential fashion,
18:34or the AC together 'cause you can do
18:37that without having dose modifications.
18:40That's why it makes sense in a
18:42three week Red room or squeeze these
18:44together in a two week regimen.
18:46In a sequential way and squeeze
18:48these together and this.
18:49This of course became the standard
18:51because the the AC from axle 2 weeks
18:53was better than AC for Taxol 3 weeks
18:55I would just emphasize that this
18:57regimen and this regimen really came
18:59out the same and so they were pulled
19:01together for the analysis and and if
19:03you can't give the cyclophosphamide
19:04with the age of my son together
19:05and if you give it the end,
19:07it would be just as effective.
19:08I have done this in certain situations with
19:10patients we're running into into issues.
19:12For example.
19:13The other thing that I've done is
19:15basically substitute Murph for adriamycin.
19:17In in this regimen,
19:18if there's issues related to
19:20cardiac toxicity,
19:20'cause we know that CMF and
19:22AC really are the same.
19:24In terms of efficacy in
19:25in the action setting,
19:27and retrospectively I'll say I think
19:29it was a shame that we did not do a
19:33comparison between AC dose, dense AC.
19:35Those dents followed by Taxol.
19:38Those stands compared to CMF.
19:40Those tents followed by Taxol
19:41'cause I bet you they come out
19:43the same and we wouldn't have all
19:45the drama about the potential for
19:47cardiac toxicity with anticyclone.
19:48Sorry I fear that we're throwing
19:50the baby out with the bathwater
19:52often when we're not using AC.
19:54Taxol,
19:54because we're afraid of cardiac toxicity
19:57and using other regimens that avoid
19:59the anti cycling and in doing so,
20:01we're also leaving out those density.
20:02I think that's a mistake and I'll show you
20:06why I think that's a mistake in a second.
20:08I just also want to emphasize and
20:09I just wanna mention this quickly.
20:11Is that the doses of the drugs
20:12we use did not come from nowhere?
20:14We actually studied the doses and
20:16we found out that moderate levels of
20:19the CF combination was equivalent
20:21to higher levels that going higher
20:23doses was not better.
20:24Half doses were inferior.
20:26This network that Dan Budman did,
20:28and in this in the cancer Community,
20:31Group B,
20:32and so the whole idea of going higher
20:34with doses to get more so killed
20:36was just not borne out by the data.
20:38The same thing was done
20:40with cyclophosphamide.
20:40And with Bernie Fisher in the NSA,
20:42BP,
20:42where they looked at higher
20:44and higher doses of
20:46cyclophosphamide in the CF combination
20:48and and it did not add in in the in in
20:52the in in the various regimens they
20:55went to higher and higher doses and they
20:58did not add so that and and Eric which
21:01I he told me he'd be listening today
21:03did the same thing with paclitaxel,
21:05going to higher dose of the 175 not
21:07showing any advantage in a study.
21:09So this notion that just going
21:11higher and higher?
21:11Doses you can get more cell kill.
21:13It's just not borne out
21:14by empirical evidence,
21:15and we have to keep that in mind as
21:17we question the original dogma that
21:19led to a lot of what we're still
21:21currently doing in in our application.
21:23Medicinal chemistry to the treatment
21:25of cancer of the present day.
21:28Well, this led to 26 randomized trials
21:32over 37,000 randomized patients looking
21:34at various permutations at dose and schedule,
21:36and this was published in The Lancet.
21:38I'm just summarizing all this work is that if
21:41you use and they talk about intensity here,
21:44but there's a very big choice of terms,
21:46but nevertheless that was the
21:47consensus that we use the term.
21:49It's really dose density,
21:51standard schedule rather than
21:53using a dose dense schedule,
21:55you get recurrences,
21:56reduced breast cancer mortality.
21:58Reduced and this is over 37,000
21:59randomized patients, so this is hard data,
22:02no.
22:02No increase in death without
22:03recurrence and there is no incremental
22:05toxicity from our agents by
22:07using them in dose dense fashion,
22:09and indeed all 'cause mortality
22:11is reduced because reducing cancer
22:13specific mortality as you see here so
22:16clearly it's shown that the concepts
22:18of those 10s therapy work and are
22:20applicable and the reason why I'm
22:22saying this is oh and by the way,
22:24is that and paclitaxel 80 weekly is superior.
22:2875 and it's a dose 10 schedule and
22:30the sideman showed this because
22:32it's being given every week.
22:33Rather than reading every three weeks
22:35and the dose response relationship
22:37for paclitaxel as Eric Weiner showed,
22:39is not is not steep and that you are
22:43accomplishing at least 1/3 as much
22:45efficacy with 80 as you are with 175.
22:49So the reason why I show all this most first
22:52of all is to catch some historical facts.
22:54For those of you who are not familiar
22:56with them but also make this point,
22:57it's gone pretty and growth is true.
23:00And you can use growth gun purchasing
23:02growth kinetics to improve cancer therapy,
23:04which leads us with the big question
23:06what is the etiology of gun?
23:07Pretty and growth?
23:08I was at something called the Ideas
23:10Festival in Aspen one year and I was
23:12having trouble parking my car so I was
23:14blocking somebody else from getting out
23:16of her parking spot and she got really
23:18angry and she came running up to me.
23:19With her hands on her hips,
23:21and I pulled down my window and she and she.
23:23And she was really angry, said,
23:24what's your problem and I said
23:26my problem is the etiology of gun
23:27protein growth what's your problem?
23:29She obviously thought I was a lunatic,
23:31which I probably am and she
23:32walked away from me.
23:33But this has been my preoccupation.
23:34For many years is understanding what
23:36is the etiology of compresion growth.
23:39And so,
23:40thinking about this in the early
23:422000s I I got a phone call from a
23:45Jean massage my great collaborator
23:47here at Morrison Kettering.
23:49He had just published his paper
23:50by Andy Mineo,
23:51was about to publish his paper by Andy Min,
23:54where they were looking at the etiology,
23:56molecular etiology and metastasis,
23:58and found this tumor.
24:00You know,
24:00which is an
24:04MMDA MB 231. Sometimes is rushing ahead,
24:07which had a certain gene
24:09expression profiling.
24:09Machines being locked, these genes
24:11being on and the tumor sticks here.
24:13But occasionally you get along with tax
24:15assist and if you get a long metastasis
24:17and you take the cells out of the lung,
24:19wash them and put them back into
24:21the memory fat pad several times,
24:23you can develop a cell line 4175,
24:26now called the lung metastasis signature,
24:28which has a signature which
24:30predicts lung metastasis because
24:32the mouse develops long itassis.
24:34He's done this for other
24:36other organs as well.
24:37Well in this paper they they had
24:39this very interesting figure.
24:40It showed that the tumor that goes
24:42to the lung more readily that has
24:44this gene expression profile also
24:46grows faster in the mammary fat pad.
24:49The one that doesn't go to
24:50the lung doesn't grow as fast,
24:51and the intermediate steps
24:54have an intermediate.
24:56Intermediate growth rate in terms of memory,
24:59fat pad.
24:59So the question that I was
25:01asked in on a phone call is.
25:04Number one, is it true?
25:05As a clinician that cancers that are
25:06metastatic tend to be faster growing?
25:08And I said that's true and he said Larry,
25:10can you figure out why and the
25:12answer was really rather obvious.
25:14Is that yes,
25:15they're getting metastatic to distant sites.
25:17Why?
25:17Why were they stop them from getting
25:20metastatic back to the original site?
25:22And he and the so the query was
25:24but the S phase fraction the KS
25:2667 was not different and I said
25:28that's makes a whole lot of sense,
25:30because basically if the tumor
25:32that goes metastatic,
25:33let's say to the lung,
25:35also gets meta meta static back to itself,
25:38then in this case we have like 3
25:40lumps that are growing independently
25:41and each of them growing at 5%.
25:43Let's say you're still going to
25:45have a growth fraction of 5%,
25:46but you're going to grow three times faster.
25:48'cause three things going at
25:505% each is going to grow faster
25:52than one thing growing at 5%.
25:53And so therefore it makes sense that
25:55they carry 67 would not be different,
25:57and yet you would get faster growth.
25:59And because it's being metastatic
26:00back to back to itself,
26:02so Jean message and I labeled
26:04this self seating and and did
26:07subsequent work in this.
26:08This was a hypothesis me on Kim.
26:11Did this work published in 2009?
26:13And this was just a brilliant experiment
26:15of the exact same tumor implanted
26:16in two different fat pads but with
26:18different fluorescent proteins in them.
26:20So they're different colors.
26:21And indeed they exchange.
26:23And this would be this.
26:24The left side of tumor.
26:25This would be the right side of tumor.
26:26This started green and and then turn
26:29red because red cells moved over.
26:31This started red and moved and
26:32and developed green.
26:33'cause green cells moved over
26:34and there's an exchange of cells
26:36between the two tumor sites.
26:37On much more work was in this paper.
26:40Obviously if you inject a non
26:42seeding tumor here and then inject
26:44the LM 2 seating tumor to the heart,
26:47it will see that tumor.
26:49Here's an interesting observation which
26:50I still think is very provocative.
26:52When you inject the tumor cells,
26:54they they light up the whole body obviously,
26:55but then over a period of 42 days
26:58they grow in the implanted tumor.
27:00That's not metastatic on this side.
27:03Why is this interesting?
27:04Because you're not developing
27:06lung metastases.
27:06In other words,
27:07the tumor is citing the the
27:09cells that you're injecting,
27:11which were developed to siedlung are
27:13not going to the lung 'cause they
27:16preferentially going to the tumor,
27:18and indeed to follow this out.
27:20If you give the tumor cells into a
27:22tail vein injection and get lung
27:24metastases first and then implanted
27:26tumor that implanted tumor will
27:28then suck cells out of the lungs.
27:30As you can see,
27:31the recipient tumor of the
27:32cells will grow at.
27:33Here 'cause it's sucking sells out
27:35a lung and these mice can actually
27:37live longer because they you can
27:39live longer with a subcutaneous tumor
27:40than you can with a long list full
27:43of metastases and and these these
27:45these have really profound implications.
27:47We think not all of which we followed up
27:49on in terms of therapeutic implications,
27:51but perhaps if we have time we
27:53can talk about them.
27:56What I want to focus in on, however,
27:58is that if cancers are growing at
28:01least partially by cells that are
28:02spreading and coming back to the tumor
28:05mass from the outside in rather than
28:07just growing from the inside out,
28:09as we always anticipated that
28:10they would grow, it would grow in
28:12this fashion like a snowflake,
28:14and this is this.
28:15This pattern of growth is with the
28:18skinny Franz is reminiscent of
28:20what the physics physicists called
28:22Diffusion limited aggregation,
28:23and it's because a water molecule.
28:25Or sell coming here is more likely to
28:27stick here than work its way into the middle.
28:29And if you do that,
28:31you actually get a pattern of growth.
28:33That's from Purtian because
28:35as objects get larger.
28:37The ratio of their surface to their
28:39volume decreases and we're going to talk
28:41more about that in in in a few minutes,
28:43and you could actually.
28:44Here's my only equation.
28:45I'm going to show you you could
28:46actually write an equation that's
28:47called the Norton mass gay equation,
28:49which basically summarizes that,
28:50and it's been much more subsequent
28:52mathematical work on this equation.
28:54And really what it means.
28:56But what it really means to me now,
28:57and I want to get into this topic.
28:59You know,
29:00with the limited time that we have,
29:02is that that this pattern of
29:04growth explains a lot of.
29:06Things that we know already about
29:08clinical medicine and not the least
29:09of which is the pattern of growth.
29:11For example, take a look at this MRI.
29:14This is a breast cancer MRI we
29:16we see this all the time and we
29:18call these satellite lesions.
29:20But frankly, it's all satellite lesions.
29:22This is a lesion.
29:23This is a delusion, this illusion.
29:24This illusion.
29:25It's got long skinny tendrils
29:26sticking out like a snowflake.
29:28It's the pattern of growth of what
29:29you'd see if the cells are coming in
29:31from the outside and so self seeding
29:33actually explains a lot about what
29:34we see in in the anatomy of cancers.
29:37At least the gross anatomy.
29:38And we'll get into the
29:40microscopic anatomy in a second,
29:41because this pattern of growth
29:44is called a fractal,
29:45and a fractal is repeated patterns
29:48at different scales and fractals
29:50have what's called a dimension.
29:52So now I'm going to go to a discussion
29:54of dimensionality because I think
29:55this is very important for some of
29:58the work that we're doing right now,
29:59and is implications particularly
30:02tumor infiltrating leukocytes.
30:03Now in Euclidean geometry.
30:05Dimensions are simple.
30:07A point has no dimension.
30:09A straight line only has length as one
30:11dimension a a sheet has two dimensions,
30:13length and and and and and
30:15and height and a cube.
30:17A solid cube has three dimensions.
30:18You're adding you're adding the depth.
30:20That's simple dimensionality
30:22in Euclidean space.
30:24In fractals it's a little
30:25bit more complicated.
30:27Let's just take one of our
30:28sheets that we had before that
30:30had a civil dimension of two,
30:32and let's look on it and cross section.
30:34Well, if you start to crumble it up,
30:35if you if you crumple up the sheet,
30:37it's going to be a little bit more than
30:39just a flat sheet and dimensionality.
30:41Here is actually 2.1 number of
30:43flat sheet is a dimension of two.
30:45If you crumple it some more it
30:47gets a higher dimension. 2.3.
30:48Now let's say that it's really getting
30:51more and more crumpled overtime.
30:53Well, it starts to. Have the appearance
30:55of something that's thicker.
30:56This is a dimension of 2.6.
30:58This is dimension of 2.8.
31:00A dimension 3 would mean you're
31:02prompted so much that it's now just
31:04a solid mass of of sheet material,
31:07but it's in a solid mass, so now it's a.
31:09It's a it's it's got.
31:10It's got the dimensionality
31:11of a 3 dimensional object or
31:13having dimensionality of three.
31:14So these are things to keep in mind and
31:17and it's a big difference between a 2.6
31:19and and and a 2.8 dimensionality you
31:21can see in terms of the thickness well.
31:24Fractals occur in nature all the time.
31:26These are artificial fractals
31:27on top of various sorts.
31:29These are the kinds of fractals
31:30that occur in nature all the time.
31:32Plants and animals and and
31:34and and diffusion in in.
31:36In substances like like ice or plastic,
31:40these these the fractals
31:42are just common in nature.
31:44Mandelbrot was discovered,
31:45there's been more Mandelbrot
31:46and written extensively about,
31:48and there's been an extensive
31:49explosion of literature in this.
31:50Written this in this regard.
31:52So what we've done is.
31:54We we've looked at this in the context
31:56of self seating and the context of
31:59leukocytes and why leukocytes because
32:01as me and Kim showed in this paper,
32:04we join mask and colleagues is an
32:07unseated state compared to a seated state.
32:09This would be an unseated tumor and
32:12this would be a tumor that's received.
32:14Received cells that have come
32:16from the outside,
32:17Ellen,
32:17two cells in the blood vessels
32:19are brought in with the seeds and
32:22they're mostly bone marrow derived
32:24endothelial cell precursors that
32:26close that blood level of growth.
32:28But I was particularly fascinated
32:29by the fact that that that when you
32:32get seating and this is is these are
32:34seated cells that they're staying green,
32:37the green for some protein.
32:38Not they have for some protein not staying,
32:41but they're obvious here in this
32:43particular setting they bring
32:44white cells in with them.
32:45CD 45 cells in with them,
32:46and so the seating process bringing
32:49bringing brings white cells in with them.
32:51Well,
32:51if it's bringing white cells in
32:53with them from the outside,
32:54perhaps the growth the pattern of
32:56white cells that we're going to see
32:58in a tumor is also going to follow,
33:00or fractal geometric pattern,
33:02and so with, with Matthew, Hannah,
33:05and and and and, and George Reese,
33:07Philo, Hannah when Ebro,
33:08G and and and and others we've looked at
33:13this by actually looking at at tumors.
33:15Conventional tumors.
33:16These are triple negative breast cancers
33:19and using image analysis in this acute pack.
33:21Roughly available image analysis
33:23program visual image analysis program
33:25to actually segment the white cells
33:27from the tumor cells so that we can
33:29actually measure the number of cells
33:31in each region of interest and then
33:33using various mathematical techniques,
33:34mathematical tricks that we've developed.
33:36We can then calculate the fractal dimension
33:38of those white cells and what we found
33:40in this is preliminary work and much
33:42more work is going on in this topic,
33:44so this is not a take home message
33:46just to show you that we've done it is.
33:48We looked at.
33:49This is the very first experiment
33:50that we did three cases of.
33:52Triple negative breast cancer
33:54and not neoadjuvant.
33:55These are patients treating
33:56the agent setting.
33:57They're small tumors versus non
33:58cases without recurrence at the
34:00fractal dimensions are different
34:02and in fact the fractal dimension
34:04of the of the white cells in the
34:06cancer that that that that recurred
34:08or that became metastatic was 2.77
34:11on the average and it was 2.65.
34:13So it's like 2.8 verse 2.6 like I
34:15showed you in previous diagram and
34:17a statistically significant P tire.
34:19Much more work is going on in this direction,
34:21but I think this is a very.
34:22Interesting area for us to think about
34:24the application of fractal geometry
34:26motivated by the concept of self
34:27seeding in terms of analyzing tills,
34:30and of course we're doing much more work
34:31in terms of characterizing those cells
34:33and and and other aspects of this work,
34:35that would be a separate talk.
34:36Hopefully at another time.
34:38What are those white cells doing in there?
34:41Well,
34:41previous work again from John Massage shop.
34:44In this one area,
34:45Korea published.
34:46This shows one of the things that they're
34:48doing is that they can actually provide
34:51resistance to chemotherapy the white cell,
34:53and this is work that's published here
34:55in cell in 2012 under stress of any sort
34:58releases a substance that causes TNF alpha.
35:02This pop up just driving me crazy.
35:05Right,
35:06right?
35:06So I'm gonna now gonna go move with the
35:08lightning speed and custom stuff out.
35:10I have no idea why that happened,
35:11but nevertheless here we are is that
35:14when you do when when we when I showed
35:16you about self seeding and when you
35:18have self seeding white cells come in.
35:20So we hypothesize that the white cells
35:22that come in the CD 45 positive cells
35:24here are coming in as a reflection of
35:26the seating process and we can uncover
35:28that by looking at their fractal geometry.
35:31And indeed we've looked at
35:32this is work with Matthew,
35:33Hannah and colleagues.
35:34We've we've done this with a.
35:36Two paths you know,
35:37method for being for segmenting
35:39between white cells and cancer cells.
35:41And indeed it is indeed fractal and
35:43the fractal dimension is different
35:44in triple negative breast cancers
35:46that recur then triple negative
35:48breast cancers that don't recur.
35:49Much more work is going on in this direction,
35:51and I discussed I described
35:52it a few minutes ago,
35:54but I can't go back over it now.
35:55We'll have to do another lecture
35:57on this particular topic.
35:58One of those white cells doing one
35:59of the things that they're doing
36:01is providing drug resistance.
36:03His work of sworn ally Acarya,
36:04and John Messages Laboratory.
36:06If you stress the cancer cells.
36:09And with anything chemotherapy or radiation,
36:13or or even heat,
36:14you can get the secretion of TNF alpha,
36:17which causes the secretion of CXCL one
36:19which goes through receptor on white cells,
36:22which causes the release of S 100
36:24proteins and can save the cancer cell
36:27as a mechanism of drug resistance.
36:29We showed this by actually showing that
36:31the inhibited by itself does nothing,
36:34but that if you give a stress
36:36in this case a chemotherapy,
36:37you can up regulate the loop.
36:39And kill cancer cells,
36:40but some are being saved by this loop.
36:42And by Ablating that loop we can get
36:44a much higher degree of cell kill.
36:46So one of the things that those
36:48infiltrating white cells is doing is
36:50providing a mechanism of drug resistance.
36:52We've also and I I told I I gave
36:56you a a wonderful anecdote here.
36:59That was a that that is lost.
37:02Now for very history about how
37:03why we did this work.
37:05But we looked at the at at those white
37:07cells that are infiltrating human cancers.
37:09And we found that very often indeed,
37:11in most cases they have leukemia
37:13genetic mutations in them.
37:15Tumor infiltrating leukocytes
37:16are not genetically normal,
37:18they are mutant,
37:19and they,
37:19however known leukemia Jennifer
37:20mutations and not only that,
37:22but if the patient is followed.
37:24In developing secondary leukemia
37:25much later in the future,
37:27those secondary leukemias have have
37:28the same mutations that you found in
37:31the tumor infiltrating leukocytes.
37:32In many cases many years earlier,
37:34there's something else that we're
37:36exploring and and doing work on on
37:38what role mutant white cells may be
37:40playing and actually and actually growth.
37:42Promotion of the cancer,
37:43as well as providing a potential
37:46mechanism for drug resistance.
37:47The last point I made in this
37:49regard or second to last point,
37:50I made this regard that you missed
37:52is that the that all of this could
37:56be exploited because circulating
37:57cancer cells in self seeding
37:59can only return to the cancer.
38:01But you can have circulating cancer cells
38:03going from one metastatic site to another,
38:05and it's been shown in both xenografts
38:07by Jonathan Weissman and also
38:09been shown in in in lung cancer.
38:12Clinical lung cancer specimens as
38:13well as some breast cancer specimens
38:15obviously can't read the details now.
38:17But this could all be exploited by
38:20giving some form of local therapy to a
38:22tumor to cause secretion of antigens,
38:24which then you can use checkpoint
38:27inhibitors and checkpoint inhibitors
38:28to get stimulation and and
38:30theoretically in this concept kill
38:32circulating cancer cells that are
38:34self seeding being drawn back to
38:36the area of inflammation that's
38:38caused by this particular
38:39procedure. We did this with with Becky
38:41Weights did this with Jim Allison when Jim
38:44Allison was at Memorial Sloan Kettering,
38:46where we looked at an animal.
38:48Model, in this case,
38:49the one that's growing in in green.
38:51If you just give anti CTA forward,
38:53nothing happens.
38:54If you just a BLT,
38:55a contralateral tumor, nothing happens,
38:58but the combination of ablation and
39:00and anti CTA 4 gets a 90% cell kill,
39:04and Heather MacArthur,
39:05who's now in Dallas has been exploiting
39:08this in a number of interesting studies.
39:11This is a work that she did
39:13at Memorial Sloan Kettering,
39:14where a primary breast tumor was ablated
39:16with crir ablation and the remaining tumor.
39:19Inside the two is profoundly
39:21Immunogen IK and we showed and
39:23published in several papers.
39:24Now in a new paper coming out of Elizabeth,
39:26Coleman has just a first authored
39:29that that you can increase the
39:32immunogenicity of that residual tumor
39:33by giving immune checkpoint inhibitors
39:35and indeed combinations work better.
39:37And this is now being looked at in
39:39terms of therapeutic implications.
39:42Coming and this could be done with
39:44radiation as well as with prior ablation
39:46which we're currently exploring.
39:47Combinations of immune checkpoint
39:49inhibitors and other thoughts
39:51related to educated T cells in
39:53terms of car T cells, for example,
39:55as well as inducing trans genes that
39:58actually may make the the the inflammation
40:00that were causing even greater.
40:02So, so this is where I left off
40:04and I just want to give you one
40:07other quick thought about geometry.
40:09You all remember that a sphere,
40:11something I mentioned to you earlier,
40:13is that the surface area is related
40:15to the square of the radius,
40:16whereas the volume is related to Cuba.
40:18The radius.
40:19This explains why mice are furry
40:21because they're very small and so they
40:23have a very high surface area related
40:25to their volume and therefore they
40:26lose heat easily and they need to be
40:28very furry to hold their heat in.
40:30You get to a large animal like an elephant.
40:32Is bald and doesn't need for her
40:35because its surface area is very
40:36low related to its volume.
40:38Its problem is getting rid of heat,
40:40which is why orphans Jen tend not to
40:41want to run very very quickly because
40:43generating heat is uncomfortable
40:44for them 'cause they don't get rid
40:46of heat very very very readily,
40:47and that is something else that we
40:50can exploit therapeutically because
40:52the fact is that as tumors grow
40:55just 'cause they're getting bigger,
40:56the ratio of their surface area is there,
40:58volume drops comes down.
41:00So you're converting basically a mouse.
41:02Into an elephant,
41:03it comes down faster for well
41:06differentiated cancers than for
41:08poorly differentiated cancers,
41:09and this is because of fractal geometry.
41:12You know if they're interested in that,
41:13we could talk about the reasons why,
41:14but that's the reason why.
41:16So that actually,
41:17if you have a tumor that's growing
41:20and you do A and the surface area
41:23decreases related to the volume
41:25while it's growing and then shrink
41:27it with chemotherapy.
41:29That the surface area to volume
41:30ratio is going to rise,
41:32and since we when when we're
41:34talking about immuno immunotherapy,
41:36we're talking about a relationship
41:37between the surface of the cancer and
41:39white cells that are trying to kill
41:41the cancer is that is that the best
41:43time to use this kind of ablation
41:45would be after an initial induction.
41:47And to take this idea and exploit
41:49it by inducing small tumor first,
41:51increasing the surface area to volume
41:53ratio and then coming in with your
41:56oblated therapy and then combining that with.
41:59Combining that with your.
42:02Your immune checkpoint inhibition.
42:05Now,
42:06the same concept can apply to in
42:09to one of the really most exciting
42:11areas in terms I think most exciting
42:13areas in terms of of modern medicinal
42:16therapy of cancer,
42:17which is the antibody drug conjugates,
42:19as we all know,
42:19they attacked a target antigen in the cancer,
42:21so they have increased payload delivery,
42:24but their penetration could be
42:25poor and this is something
42:27that has to be exploited when
42:29they're internalized that the the
42:30payload is reduced, it is is is
42:32released in 'cause the cancer cell.
42:34But more than that.
42:35In terms of the activity of these
42:37payloads on killing the cancer cell,
42:39they often leak out and they can kill
42:42adjacent cells that don't necessarily
42:43have that particular target.
42:45This or this work of Josh Drago.
42:47Well, this can be exploited by the same
42:50way is that when you if you used your
42:53antibody drug conjugate to a large tumor,
42:55you get down regulation of the target,
42:57and that's not not what you
42:59want to optimize the effect.
43:01So one of the things we're exploring,
43:02and this is not in the clinic yet.
43:04This is just an experiment experiment
43:05that we're doing right now,
43:06preclinical in preparation for clinical
43:08experiment is by giving a non ADC induction.
43:12First we can increase the
43:13surface area to volume ratio.
43:15And then come in with the ADC as a
43:17late intensification and therefore
43:18it should be even more active in this
43:21area to get tumor volume eradication.
43:23And if the animal experiments work,
43:25I think there's something else that
43:27could be exploited extremely easily
43:28in the clinic 'cause we have a lot
43:29of drugs in breast cancer that can
43:31cause tumor volume regression that
43:32are not Adcs and then instead of
43:34waiting for the tumor to grow and
43:36using using your Adcs in as a salvage
43:38if you use them at time of maximum
43:40tumor volume regression and this,
43:42by the way,
43:43could be determined not just by actually
43:44watching the cancer shrink with imaging.
43:46But also by by the burden of
43:48of circulating cancer and DNA,
43:50which would be another way of
43:52actually when that plateaus,
43:53you know you've achieved your maximum
43:55volume regression would be the best
43:56time to come in with your ABC's.
43:59Last slide and I'm not going
44:01to talk about this, obviously,
44:02is that we're exploiting exploiting
44:04all of this in much more sophisticated
44:07mathematics with a number of
44:09mathematical collaborators.
44:10I don't album and Jodeci in particular.
44:13Arena Elkin and jungle in terms
44:15of actually looking at this same
44:17mathematical concepts in terms of gene
44:20gene interactions and their networks.
44:22The same thing that works at the cell
44:23level and the the tumor of brain
44:25leukocyte level may work at the gene level.
44:27This would have a different fractal
44:28dimension than this, for example.
44:29'cause we have a lower fractal dimension.
44:31This would have a higher fractal dimension.
44:33You can look at gene networks in
44:34the same way as another term for
44:37this cord curvature.
44:38Obviously I can't get into it,
44:39but this is giving us some great
44:41insights and we recently published a
44:43paper in ovarian cancer that actually
44:45showed that the the structure of the
44:47gene gene interaction network has
44:49predicted values in terms of response
44:51to immune checkpoint inhibition.
44:53In this situation and,
44:54and indeed that you can actually
44:56predict which patients with ovarian
44:57cancer there's not supposed to
44:59respond to immune checkpoint ambition.
45:01Will respond on the basis of the the
45:04mathematical analysis of of their.
45:06Gene Gene interactive networks.
45:08So what I've been able to do,
45:10I hope in this lightning talk made even
45:12more lightning by the loss of the Internet.
45:17It's just described where this all came from.
45:19Skippers model being modified to the
45:22compression growth model and leading to
45:26a clinical advance and then why tumors
45:29grow in that kind protein fashion.
45:31The whole self seating concept which led
45:33us into the concept of fractal geometry,
45:36which is now one of my most active areas.
45:38Investigation how,
45:39how can we actually quantify tills and
45:41what is the prognostic significance
45:43of them using fractal geometry?
45:45How does all of this relate
45:46to drug resistance?
45:47And optimizing immunotherapy and optimizing
45:49new agents such as antibody drug conjugates.
45:52Forgive me for speaking too fast,
45:54but I I know we have to end on time
45:55and thank you all for listening.
45:57I apologize that we lost the Internet.
46:00Thank you Larry. That
46:01was if we have a couple of minutes let
46:02you know if I can do a couple of talks.
46:04I can stay later if people want to
46:06stay late. We have a couple of questions.
46:08First of all, thank you so much for you.
46:10Know it just to me. It's amazing for
46:13a Conservatory trained musician to be
46:16such mathematician at the same time,
46:18I don't know how both sides of the border
46:20link they get. They link together music
46:21and math is the same, the same the same.
46:23You know part of the brain. So
46:25we have some questions from Pat Larussa
46:28and David Rim to start with there.
46:30Kind of on the same pattern on
46:32the fractal pattern differences
46:33between hormone receptor positive
46:35and triple negative breast cancer.
46:38Are there differences that you see
46:39not only for the tumor, but the tills?
46:41And then does that work in
46:43terms of the agency used
46:45colleagues?
46:45That's what we work in progress,
46:47but but the answer is almost
46:49certainly so because you know,
46:50I'm not doing anything with the
46:52fractal geometry that the pathologist,
46:54they scope pathologist is doing
46:55with their eyes. You know,
46:57a skilled pathologist looking and says,
46:58hey, this is well differentiated.
46:59This poor differentiated differentiation.
47:01Is poor differentiated means
47:03a high fractal dimension,
47:05whereas a well differentiated
47:06means a low fractal dimension?
47:08And so basically I'm just
47:10basically just quantifying.
47:11I'm quantifying something that that the eyes
47:13of the beholder have seen is seen already.
47:15So clearly we're gonna see this.
47:17But you're talking here about fractional
47:19dimension of the cancer cells,
47:20which is obviously something we're exploring.
47:22I was talking about fractal
47:23dimension of of of the tills,
47:25but it all relates together and and I think
47:27what makes it really intriguing to me.
47:30Just personally,
47:30maybe nobody else, but to me.
47:32Is that it relates to this
47:33concept of a pattern of growth?
47:35This self seeding pattern of growth?
47:37One thing you gotta know about math is that
47:39you know even if self seeding didn't happen.
47:42If if things anatomically look the way
47:44they would happen were it to happen,
47:47it still is biologically significant.
47:49That's that's the way that's the
47:51way math mathematics works alright.
47:52You don't have to have the
47:53the example you know the the.
47:55The same mathematics works for
47:57gravity and for magnetism,
47:58even though the mechanisms are different,
48:00we don't understand the mechanisms,
48:01but we know they're different at the same,
48:03the same, the same.
48:04The same.
48:04You know the same mathematics
48:05you know works for the theory of
48:07universal gravitation works the same.
48:08So so once we actually understand
48:10the mathematical principles,
48:11they can generalize even if the
48:12thing that got us into that which is
48:14substituting concept is not valid.
48:16But I really do think the self
48:18seeding thing is valid.
48:19But based on the on the accumulating
48:21body of evidence that we're seeing,
48:22so I'm just basically trying
48:23to quantify that.
48:25Thank you we have another question
48:27on the implications of Gumpertz and
48:29growth for the rate of survival
48:31and proliferation of cancer cells.
48:32What are there implications and then
48:35does it imply that proliferation slows?
48:37And if so, why are the clinical
48:40implications of that slowed growth?
48:42All
48:42right? You know, you're asking for
48:44treatise and a little bit comma,
48:45and I wrote a really nice review
48:47article about the clinical implications
48:49of cancer self seating so you know,
48:51COMEN and Norton.
48:52You can Google it and go with that paper.
48:54Really, really, really very quickly.
48:55When we go into all that in great depth,
48:57first of all,
48:58gun person growth has to happen.
48:59'cause if it didn't happen we we
49:01would have no chance against cancer
49:03because with exponential growth I
49:05mean from the time of diagnosis,
49:06time of death would be a matter of of of
49:09of weeks at even even for solid tumors.
49:11So we know there's gotta be a tailing
49:13off of growth rates and it really
49:15has great profound implications
49:16in terms of our understanding,
49:18growth and and and planning for therapy.
49:20I I think it's a shame that we haven't
49:22used those dense sequential therapy
49:24for more tumors beside breast cancer.
49:25There's been a little bit of work
49:27in lymphomas in this regard.
49:28A little bit of work in other tumors,
49:29but we haven't optimally exploited it,
49:31and I think that we could actually
49:32do better even with existing agents
49:33if we were able to take some of the
49:35principles we learned with breast cancer,
49:37move them into that setting.
49:38But right now where I'm focusing
49:40in on instead,
49:41is how do we use some of the newer agents,
49:43particularly Adcs,
49:44and apply some of the things we've
49:46learned from chemotherapy to it
49:48using gun protein growth and using
49:50our concepts with tumor geometry.
49:53And maybe the last question is from
49:57Doctor Bafan. Thoracic surgery?
49:59Is the self seating limited to cancer
50:02cells or do other employee put in stem
50:05cells from normal cellular turnover,
50:07preferentially land and tumors,
50:09for example to gastro intestinal stem cells?
50:14Go on to blood, still some lines and other.
50:17Yeah, it's a great question.
50:18It's a great question because it's
50:19something that that that we are on verbal.
50:21Yeah, stem cells and seeds I think
50:23is the same thing. Basically,
50:24I think that's the capacity of stem cells
50:26is being able to move around and and.
50:28And frankly it's not such a stretch.
50:30'cause that's what happens in Embryology.
50:31I mean, that's that's how
50:32the embryo forms is,
50:33that is that the the stem cells move
50:35from one spot to another in a very,
50:37very logical kind of fashion.
50:39It isn't that and people
50:40always ask that you know,
50:41you know what draws them self to that site.
50:43It isn't drawn to that site.
50:45We know this from this from the self seeding
50:46work that's been done in the laboratory.
50:48The cells go all over, it's just
50:50where they stick that really matters.
50:51So it looks like it's drawn to that
50:52site only 'cause they stuck there and
50:54and it's that sticking their stickiness
50:55that I think is something that that
50:57that that's being scored by a number
50:59of by a number of investigators.
51:02You know that particular phenomenon,
51:05but I'm sure this happens in general.
51:06Look at wound healing.
51:07I mean wound healing.
51:08You know you heal your wound,
51:09your surgeons,
51:09you don't heal your wound because of the
51:11cells that are right there where you cut you.
51:13The cells are brought in there,
51:14you know,
51:14married to rod cells are brought
51:15in there and that's what allows.
51:16The wound to heal so that so that I I
51:18think seating is a general biological
51:20phenomena and a lot of things that we're
51:22doing in cancer may relate to other things,
51:24such As for instance, wound healing.
51:25Uhm, uh,
51:26that that we're starting to think you know,
51:29you know about the cytokine release
51:31syndrome that we're seeing with COVID-19,
51:33and how that relates to the mobility of
51:34of of white cells in that regard as well.
51:36In response to inflammation.
51:37So so it may be a much
51:39more general phenomenon.
51:40The cool thing for me,
51:41and again,
51:42I'm just speaking for me,
51:42is that the mathematics we
51:44workout in one area may relate
51:45to all these other areas as well.
51:46And that once we understand,
51:48developed these mathematical principles,
51:49that we can actually use them to generalize
51:52beyond cancer into heart disease.
51:53We know that colonial meta polices cells
51:56are are important in arteriosclerotic
51:57heart disease as well as as as we
52:00just discussed with cancer as well,
52:02so that these principles may generalize
52:04and and and have much more replicability.
52:07Can we squeeze one more question.
52:08This is from an Chang former
52:11memorial colleague who are now
52:12our Chief Network Officer.
52:14She's asked, can we exploit Atascosa
52:16specific or other gene processes in
52:19the tumor microenvironment to prevent?
52:21Self seating in the niche of growth.
52:23Yeah yeah, great another great question.
52:25Another great hot topic.
52:26You know something that Joe and I,
52:28Joe and Megan.
52:29I thought a very early days when we
52:30started doing this when we started
52:32doing this work and I remember that
52:34we published the paper with 2009,
52:35so it's been a lot of time this
52:37past and and and and you know,
52:38we know that that that cytokines in
52:40flammatory cytokines are important
52:41for the process and that's already
52:43that's already been demonstrated and
52:45that may be why inflammation is is
52:46such a problem is related to cancer.
52:48But I want to get the things that
52:50are more targetable than that.
52:51And so that's one of the reasons
52:53why that very last slide that I
52:55showed you that very complicated
52:56mathematical slide is is is we we,
52:58we are right now doing a number of
53:00different studies looking at gene
53:02interactive networks using the
53:04same basic mathematical principles.
53:06In fact,
53:06trying to see what are the gene
53:08interactions that may may underlie that
53:10process, because that will tell us what,
53:12what,
53:12what genes we maybe have
53:14development chemicals to,
53:15medicines to to be able to be able to target,
53:17to interfere with this,
53:18the just there's something in that
53:20regard I think is is really important.
53:22Is that?
53:22We focused on so much of our energy
53:24in terms of medicinal chemistry on
53:26targeting genes or gene products,
53:28and one of the things we're learning by
53:29using that mathematics and looking at
53:31gene interaction networks is this is yes,
53:33indeed is the action of individual genes,
53:35but it's not the action of
53:37individual genes by themselves.
53:37They're all interacting with each other,
53:39and it's the whole network of
53:41genes that actually forms a a
53:43meaningful biological entity and
53:44not just the individual genes.
53:46So we're going to have to target
53:48target those interactions rather
53:50than target the genes themselves,
53:51and that's not something that we commonly.
53:53Do you?
53:54Although we we probably do it,
53:56we don't realize we do it with
53:57with with therapy.
53:58When you give steroids to a patient
54:00for all the reasons that we give
54:02Google Corticoids for a patient you
54:04you're attaching to Google Corticoid
54:05receptors all over the place,
54:07not just in a particular place,
54:09and you're basically affecting
54:10cell cell interactions all over
54:11the place and you're affecting
54:13gene interaction networks all over
54:14the place by using some of the
54:16most powerful drugs that we have
54:18actually are not targeted therapy,
54:20it's starting to question the notion of
54:22are we really better off using targeted
54:24therapy when we're dealing with complex?
54:26Processes or should we be able to
54:29target the complexity itself so so
54:31that's one of the things that we're
54:33zeroing in on on that particular thing.
54:34Now,
54:35how do we find those drugs is is that?
54:37Basically,
54:37if you understand the networks and you can,
54:40you could then screen a lot of
54:42different drugs and see how it
54:44affects the network and so you can
54:45actually as as possible that even old
54:47drugs could be repurposed for this reason.
54:49And you may not be able to put your
54:51finger on exactly why they work,
54:52but you could just show that
54:53they are working in the show.
54:54They have clinical utility.
54:56And and and that's that's really
54:58kind of a very different way of
55:00thinking about medicinal chemistry.
55:01Rather than saying I,
55:02I'm gonna go after the specific
55:04target to actually go after
55:06basically the biological effect.
55:07Or you know,
55:08in in general with your agents
55:09and then and then move them into
55:11clinic on that kind of basis.
55:12So those are some of the things that
55:13we're thinking about right now.
55:15Thank you Larry.
55:15This is been really great and
55:17we really appreciate your time.
55:18I know next year Eric will want
55:20to have you here in person again
55:23to talk to us and this was really
55:26just phenomenal lecture.
55:27Even though you dropped
55:28off for a few minutes,
55:29if you were able to bring everything back
55:31and and and please thank thank
55:32the person who actually called me
55:33on my cell phone so that I they
55:35got to me so I was able to come
55:37back in I I appreciate it.
55:38Thank you all very much for
55:39listening. Thank you Larry.