Tracking Cancer Cells

Transcript

00:00 --> 00:03Funding for Yale Cancer Answers is
00:03 --> 00:05provided by Smilow Cancer Hospital.
00:05 --> 00:08Welcome to Yale Cancer Answers
00:08 --> 00:09with doctor Anees Chagpar.
00:09 --> 00:11Yale Cancer Answers features the
00:11 --> 00:13latest information on cancer care
00:13 --> 00:15by welcoming oncologists and
00:15 --> 00:17specialists who are on the forefront
00:17 --> 00:19of the battle to fight cancer.
00:19 --> 00:21This week it's a conversation about new
00:21 --> 00:23research into cell mutations and cancer
00:23 --> 00:26therapies with Doctor Jeffrey Townsend.
00:26 --> 00:28Dr. Townsend is the Elihu Professor
00:28 --> 00:30of Biostatistics and professor of
00:30 --> 00:32ecology and evolutionary biology
00:32 --> 00:34at the Yale School of Medicine,
00:34 --> 00:37where Doctor Chagpar is a professor
00:37 --> 00:38of surgical oncology.
00:39 --> 00:40So maybe we can start off, Jeff,
00:40 --> 00:42by you telling us a little bit more
00:42 --> 00:44about yourself and what it is you do.
00:45 --> 00:48I'm in the Biostatistics department at Yale,
00:48 --> 00:51but I'm perhaps the most biological
00:51 --> 00:52of the members of the department
00:52 --> 00:55in that all my degrees are biology
00:55 --> 00:58and what I work on is large scale
00:58 --> 01:01genomic data sets about the genomic
01:01 --> 01:04mutations that change tumors and
01:04 --> 01:08what leads to tumors and also the
01:08 --> 01:10exogenous and endogenous factors
01:10 --> 01:12that make us come down with cancer.
01:13 --> 01:15So let's dive into
01:15 --> 01:17that a little bit more.
01:17 --> 01:19Many of our listeners may know
01:19 --> 01:21about what the genome is.
01:21 --> 01:23Basically the conglomeration
01:23 --> 01:27of DNA that makes us who we are.
01:27 --> 01:29But tell us a little bit more
01:29 --> 01:31about genomics and the
01:31 --> 01:33study of these mutations.
01:33 --> 01:35Yeah, I think the thing that's
01:35 --> 01:37important to understand about the work
01:37 --> 01:38that we do is that we're working on
01:38 --> 01:40what are called somatic mutations.
01:40 --> 01:42So not what you inherited from
01:42 --> 01:44your mother or from your father,
01:44 --> 01:45but rather the mutations that
01:45 --> 01:48occur in your body during the
01:48 --> 01:49time that you're developing.
01:49 --> 01:51These are the kinds of mutations
01:51 --> 01:53that people talk about trying to
01:53 --> 01:55avoid by not smoking or not being
01:55 --> 01:56exposed to too much UV light.
01:56 --> 01:59So we look at those kinds of mutations
01:59 --> 02:01that accumulate during your lifetime
02:01 --> 02:03and then lead to cancer on top of
02:03 --> 02:05all the germline variation that
02:05 --> 02:07you have coming from your parents.
02:08 --> 02:10And so tell us more about kind of
02:10 --> 02:13how that works and how you
02:13 --> 02:15discover these mutations and
02:15 --> 02:18how you actually define that these
02:18 --> 02:20particular mutations have an impact
02:20 --> 02:23in terms of cancer generation.
02:23 --> 02:26This is a really, really important
02:26 --> 02:28topic since the human genome,
02:28 --> 02:30we've developed lots of technologies
02:30 --> 02:32that allow us to sequence genomes,
02:32 --> 02:34including the genomes of tumor tissue
02:34 --> 02:36as opposed to your normal tissue.
02:36 --> 02:39And by comparing that tumor tissue sequence
02:39 --> 02:42to the sequence we see from your blood
02:42 --> 02:44or from some normal adjacent tissue,
02:44 --> 02:47we can uncover all the genetic mutations
02:47 --> 02:49that are specific to the tumor and
02:49 --> 02:52aren't natural to the rest of your body.
02:52 --> 02:56And those mutations tend to be of two kinds.
02:56 --> 02:58Some are just mutations that just happen
02:58 --> 03:00to have happened and don't really lead
03:00 --> 03:03to cancer and other ones lead to cancer.
03:03 --> 03:05And so differentiating between those
03:05 --> 03:08two can be done by just looking at the
03:08 --> 03:11frequencies that certain mutations occur
03:11 --> 03:13and understanding what the underlying
03:13 --> 03:16rate at which those mutations occur are.
03:16 --> 03:18And by combining those two factors together,
03:19 --> 03:21we can get a quantitative estimate of
03:21 --> 03:24exactly how much of the cancer is being
03:24 --> 03:27caused by particular mutations in the genome.
03:27 --> 03:29So I'm very excited about research
03:29 --> 03:31we're doing that allows us to take that
03:31 --> 03:33quantitative estimate and do things.
03:33 --> 03:34This is very preliminary at this point,
03:34 --> 03:37but like actually assess in an individual
03:37 --> 03:39why did that person get their cancer.
03:39 --> 03:40And it's not just saying, Oh,
03:40 --> 03:41well,
03:41 --> 03:43we know smoking causes cancer or we
03:43 --> 03:45know UV light can cause Melanoma,
03:45 --> 03:46but actually looking at the individual
03:46 --> 03:48and saying, in your case,
03:48 --> 03:50why did that cancer arise?
03:51 --> 03:53So tell us more about the study
03:53 --> 03:56itself because I can imagine that
03:56 --> 04:00as you look at these mutations and
04:00 --> 04:02when you compare tumor DNA to
04:02 --> 04:05normal DNA that there are more mutations.
04:05 --> 04:07You might be able to say, OK
04:07 --> 04:10there are more mutations and
04:10 --> 04:12hypothesize that those more mutations
04:12 --> 04:15are what actually caused the cancer.
04:15 --> 04:18But causation and association are different.
04:18 --> 04:21So how did you establish that?
04:21 --> 04:24And what is the long term impact of
04:24 --> 04:27being able to define in a particular
04:27 --> 04:29individual what mutations cause their
04:29 --> 04:32cancer? Because once they have cancer,
04:32 --> 04:34isn't it kind of like a fait accompli, or
04:37 --> 04:40does defining those mutations actually
04:40 --> 04:43have an impact then on how they're treated?
04:43 --> 04:46Yeah. So first, in terms of defining
04:46 --> 04:49which mutations are leading to cancer,
04:49 --> 04:50what's important to understand
04:50 --> 04:52is that there are very,
04:52 --> 04:54very different rates of mutation for
04:54 --> 04:56different sites within your genome.
04:56 --> 04:59Some parts of the genome are much
04:59 --> 05:01more exposed just because of the
05:01 --> 05:02way the genome replicates and
05:02 --> 05:04things like that, than others.
05:04 --> 05:06And so what we have to do is
05:06 --> 05:07estimate how likely every site
05:07 --> 05:10in the genome is to be mutated.
05:10 --> 05:11And then from that look at
05:11 --> 05:12the tumors and say,
05:12 --> 05:14do we see that frequency of mutation in
05:14 --> 05:17the tumor or do we see it more often?
05:17 --> 05:19And if it's more often,
05:19 --> 05:21then it must be leading to tumors
05:21 --> 05:22because that's what we're sequencing and
05:22 --> 05:25seeing more of it there than we would expect.
05:25 --> 05:27So that's how we differentiate
05:27 --> 05:28those ones that are causing
05:28 --> 05:30cancer from those that aren't.
05:30 --> 05:32And then why is it important?
05:32 --> 05:32Well,
05:32 --> 05:34in addition to sort of the question
05:34 --> 05:36that I introduced originally,
05:36 --> 05:37like trying to understand
05:37 --> 05:39why individuals get cancer.
05:39 --> 05:41Cancer continues to evolve.
05:41 --> 05:44It's not just a
05:44 --> 05:45static thing that you have,
05:45 --> 05:47but it actually changes
05:47 --> 05:49over time in individuals.
05:49 --> 05:53If you come in and you have a tumor removed,
05:53 --> 05:54but you have recurrence,
05:54 --> 05:56then one of the things that the
05:56 --> 05:58physicians are charged with doing is
05:58 --> 06:00trying to understand why that recurrence
06:00 --> 06:02occurred and trying to negotiate
06:02 --> 06:04around the evolution of that tumor.
06:04 --> 06:06So the tools that have just been
06:06 --> 06:09released now that we've been using
06:09 --> 06:11allow one to understand what
06:11 --> 06:13that trajectory of evolution is.
06:13 --> 06:14In other words,
06:14 --> 06:16you're at this state right now genetically,
06:16 --> 06:18but what's the next genetic change
06:18 --> 06:20likely to be in a probabilistic
06:20 --> 06:22way and what's the next one
06:22 --> 06:24after that likely to be?
06:24 --> 06:25And it's the first time we've
06:25 --> 06:27really been able to
06:27 --> 06:29characterize that in terms of a
06:29 --> 06:31trajectory of change where we
06:31 --> 06:32understand quantitatively how much
06:32 --> 06:34each mutation is increasing the
06:34 --> 06:37survival and proliferation of these cells.
06:38 --> 06:39So, that's interesting.
06:39 --> 06:42How exactly do you do
06:42 --> 06:45that in terms of defining, OK,
06:45 --> 06:49this mutation caused your cancer and
06:49 --> 06:53probabilistically you have an X percent
06:53 --> 06:55probability of getting a recurrence.
06:55 --> 06:58Tell us more about if that's really
06:58 --> 07:01what you can do in an individual
07:01 --> 07:03way and how exactly you
07:03 --> 07:06come up with that probability.
07:06 --> 07:08Right. So what we work with,
07:08 --> 07:10as typical with these kinds
07:10 --> 07:11of studies, is population.
07:11 --> 07:12So we don't know necessarily
07:12 --> 07:14for an individual what their
07:14 --> 07:15next change is going to be.
07:15 --> 07:18But what we can do is look at lots of tumors,
07:18 --> 07:19see which changes have occurred
07:19 --> 07:22and in some sense order them in
07:22 --> 07:24individual tumors and then say, oh,
07:24 --> 07:26given where you are on this trajectory,
07:26 --> 07:29what's the next mutation likely to be?
07:31 --> 07:34And so I can imagine that for
07:34 --> 07:37people who may be listening,
07:37 --> 07:38they may be saying to themselves,
07:38 --> 07:40well, that's great. You know,
07:40 --> 07:43you can give me an estimate of what
07:43 --> 07:45my next mutation is going to be.
07:45 --> 07:47Has there been work to kind of say, well,
07:47 --> 07:50how do we prevent that from happening?
07:50 --> 07:52How do we prevent your next recurrence?
07:52 --> 07:54Yeah, that's what we're working on
07:54 --> 07:56right now with this approach is to
07:56 --> 07:58better understand and better line up
07:58 --> 08:00essentially what we know about these
08:00 --> 08:02genetic changes and how they occur,
08:02 --> 08:04what order they occur with the
08:04 --> 08:06kinds of precision medicines that
08:06 --> 08:08are now being developed at a more
08:08 --> 08:10breakneck pace through
08:10 --> 08:12the great research that's
08:12 --> 08:14happening here at Yale and elsewhere.
08:14 --> 08:17And the point is that all of those
08:17 --> 08:19different precision treatments can
08:19 --> 08:21be marshaled in different ways.
08:21 --> 08:22And it's getting more and more complex
08:22 --> 08:25to sort of think through how to treat
08:25 --> 08:27an individual when they have this sort
08:27 --> 08:29of evolving cancer that is evolving
08:29 --> 08:30resistance to different therapies.
08:30 --> 08:34And so hopefully what we can do with
08:34 --> 08:36our genetic trajectories is to inform
08:36 --> 08:39for a patient's decision making and
08:39 --> 08:41for a physician's decision making
08:41 --> 08:43about the next therapy that must be
08:43 --> 08:46prescribed to someone who has cancer.
08:46 --> 08:48What would be the best trajectory to
08:48 --> 08:51occupy in terms of the genetic evolution?
08:51 --> 08:53And are there ways we can corner the
08:53 --> 08:55cancer essentially so it can't evolve
08:55 --> 08:57resistance and lead to a recurrence?
08:59 --> 09:03So kind of trying to treat to the
09:03 --> 09:06cancer currently in a way that they
09:06 --> 09:09then don't mutate according to the
09:09 --> 09:11trajectory that you've hypothesized
09:11 --> 09:12that they otherwise would?
09:13 --> 09:14That's exactly right.
09:14 --> 09:16Let me give you an example from
09:16 --> 09:18other work that we did which
09:18 --> 09:21was looking at EGFR therapy.
09:21 --> 09:23This is an irlatinib therapy that is
09:23 --> 09:25not actually given currently,
09:25 --> 09:28but when we looked at
09:30 --> 09:32irlatinib therapy,
09:32 --> 09:34one of the things that we noticed
09:34 --> 09:36was that cisplatin therapy which
09:36 --> 09:38is often given in the context
09:38 --> 09:40of EGFR driven lung cancer can
09:40 --> 09:42actually lead to the underlying
09:42 --> 09:45mutations that give you resistance,
09:45 --> 09:47give the tumor resistance
09:47 --> 09:48to erlatinib therapy.
09:49 --> 09:51So that's an example where you
09:51 --> 09:52wouldn't want to order the
09:52 --> 09:54particular treatments cisplatin
09:54 --> 09:56and then erlatinib because you're
09:56 --> 09:57basically creating the genetic
09:57 --> 10:00variation in the tumor so that
10:00 --> 10:01it can evolve resistance very
10:01 --> 10:04quickly once you give the therapy.
10:05 --> 10:09And so as we think about
10:09 --> 10:12the idea that you may be able to
10:12 --> 10:15understand better how tumors evolve
10:15 --> 10:18in terms of their genetic mutations
10:18 --> 10:22which can kind of bypass some of our
10:22 --> 10:24therapies and cause resistance.
10:24 --> 10:27One can only think about how do
10:27 --> 10:30you take this into the preventative arena.
10:30 --> 10:32So if we know, for example,
10:32 --> 10:35that UV light causes certain mutations
10:35 --> 10:38or smoking causes certain mutations,
10:38 --> 10:41is there a way to use the information
10:41 --> 10:44that you have been able to garner
10:44 --> 10:46so far to think about whether there
10:46 --> 10:48are preventative treatments that
10:48 --> 10:51can actually stop the mutations from
10:51 --> 10:53occurring in the 1st place that
10:53 --> 10:55give people their initial cancers?
10:56 --> 10:58As a member of the School of Public Health as
10:58 --> 10:59well as a member of Yale Cancer Center,
10:59 --> 11:01I think about prevention a lot.
11:01 --> 11:04And one of the things that I'm really
11:04 --> 11:07hopeful we can do is to use the methods
11:07 --> 11:09that we've designed to better characterize
11:09 --> 11:12what has led to cancer in individual cases,
11:12 --> 11:15to give more information to patients
11:15 --> 11:18so that they can share it with their loved
11:18 --> 11:20ones and the ones that they care about.
11:20 --> 11:22And so if they find out
11:22 --> 11:23that their cancer was, say,
11:23 --> 11:25caused by smoking or caused
11:25 --> 11:26by UV light exposure,
11:26 --> 11:29those individuals who are close to them
11:29 --> 11:31can know some of these risk factors
11:31 --> 11:34that affected them and led to their cancer.
11:34 --> 11:36And hopefully that kind of peer education
11:36 --> 11:39I think can play a role in helping to
11:39 --> 11:42prevent many of these exogenous factors,
11:42 --> 11:44these factors outside of the body
11:44 --> 11:46that can lead to cancer.
11:47 --> 11:49Yeah,
11:49 --> 11:53I hope that most people know that
11:53 --> 11:55smoking leads to cancer and UV
11:55 --> 11:58light leads to cancer and there
11:58 --> 12:00is good public awareness of that.
12:00 --> 12:03What I'm kind of thinking about is
12:03 --> 12:07if you're doing work that looks at
12:07 --> 12:10how these exogenous factors can
12:10 --> 12:13actually change the genomic profile
12:13 --> 12:16that causes cancers and understand
12:16 --> 12:18the trajectory by which those
12:18 --> 12:20cancers have further mutations,
12:20 --> 12:22is there a way to prevent
12:22 --> 12:23the initial mutation?
12:23 --> 12:25So for example,
12:25 --> 12:28one could imagine that just like
12:28 --> 12:31you have drugs that can kind of
12:31 --> 12:34direct cancers into either for
12:34 --> 12:37causing more resistance or less
12:37 --> 12:39resistance to further therapies
12:39 --> 12:42and there are further mutations.
12:42 --> 12:44I could imagine that you could have,
12:44 --> 12:46you know what a sunscreen that
12:46 --> 12:50would prevent the UV light from
12:50 --> 12:53causing certain mutations or an
12:53 --> 12:57inhaler that might prevent cigarette
12:57 --> 13:00smoke from causing mutations.
13:00 --> 13:04Just a thought to kind of think about,
13:04 --> 13:06but we have to take a quick
13:06 --> 13:08break for a medical minute,
13:08 --> 13:09so we'll pick up the
13:09 --> 13:10conversation right after that.
13:10 --> 13:13Please stay tuned to learn more about
13:13 --> 13:15tracking cancer cells with my guest,
13:15 --> 13:16Doctor Jeffrey Townsend.
13:17 --> 13:19Funding for Yale Cancer Answers comes
13:19 --> 13:21from Smilow Cancer Hospital where
13:21 --> 13:23the lung cancer screening program
13:23 --> 13:26provides screening to those at risk
13:26 --> 13:27for lung cancer and individualized
13:28 --> 13:30state-of-the-art evaluation of lung nodules.
13:30 --> 13:35To learn more, visit smilowcancerhospital.org.
13:35 --> 13:37Genetic testing can be useful for
13:37 --> 13:39people with certain types of cancer
13:39 --> 13:41that seem to run in their families.
13:41 --> 13:43Genetic counseling is a process
13:43 --> 13:45that includes collecting a detailed
13:45 --> 13:46personal and family history,
13:46 --> 13:48a risk assessment,
13:48 --> 13:51and a discussion of genetic testing options.
13:51 --> 13:53Only about 5 to 10% of all cancers
13:53 --> 13:55are inherited and genetic testing
13:55 --> 13:57is not recommended for everyone.
13:57 --> 14:00Individuals who have a personal and
14:00 --> 14:02or family history that includes
14:02 --> 14:04cancer at unusually early ages,
14:04 --> 14:06multiple relatives on the same side
14:06 --> 14:08of the family with the same cancer,
14:08 --> 14:11more than one diagnosis of cancer in
14:11 --> 14:13the same individual, rare cancers,
14:13 --> 14:16or family history of a known altered
14:16 --> 14:18cancer predisposing gene could be
14:18 --> 14:20candidates for genetic testing.
14:20 --> 14:22Resources for genetic counseling and
14:22 --> 14:25testing are available at federally
14:25 --> 14:26designated comprehensive cancer
14:26 --> 14:28centers such as Yale Cancer Center
14:28 --> 14:30and Smilow Cancer Hospital.
14:30 --> 14:33More information is available
14:33 --> 14:34at yalecancercenter.org.
14:34 --> 14:36You're listening to Connecticut Public Radio.
14:37 --> 14:39Welcome back to Yale Cancer Answers.
14:39 --> 14:41This is Doctor Anees Chagpar
14:41 --> 14:43and I'm joined tonight by my guest,
14:43 --> 14:44Doctor Jeffrey Townsend.
14:44 --> 14:47We're talking about his work looking
14:47 --> 14:50at mutations and how these mutations
14:50 --> 14:53can influence each other in a way
14:53 --> 14:55that affects cancer evolution.
14:55 --> 14:58And to that end, you know, Jeff,
14:58 --> 15:00maybe you can talk a little bit
15:00 --> 15:02more about the actual techniques of
15:02 --> 15:03the work that you've been doing.
15:03 --> 15:05And you know,
15:05 --> 15:07whether it's that you have found
15:07 --> 15:10that there's just one mutation that
15:10 --> 15:13occurs that kind of leads to a
15:13 --> 15:16series of steps that then cause
15:16 --> 15:18cancer and recurrence or whether
15:18 --> 15:20there's actually multiple mutations.
15:20 --> 15:22And if you only have one,
15:22 --> 15:24it may not lead to anything.
15:24 --> 15:27And so maybe disrupting the
15:27 --> 15:29interactions between these mutations
15:29 --> 15:31actually has a role to play.
15:31 --> 15:33Can you can you talk a little
15:33 --> 15:33bit more about that?
15:34 --> 15:36Absolutely. Let me give a little
15:36 --> 15:38bit of context, which is over the
15:38 --> 15:41past decade or even a little more,
15:41 --> 15:42there's been a very,
15:42 --> 15:44very concentrated effort
15:44 --> 15:47to find these mutations that underlie cancer.
15:47 --> 15:48And many groups are doing it,
15:48 --> 15:50not just mine, of course.
15:50 --> 15:53And and have been for many years now,
15:53 --> 15:54as I said, almost a decade.
15:54 --> 15:57So that effort has largely focused
15:57 --> 15:59on the identification or the
15:59 --> 16:01discovery of gene naming, oh,
16:01 --> 16:04this gene is actually relevant to cancer,
16:04 --> 16:05or that gene is relevant to cancer,
16:05 --> 16:07or this gene is not.
16:07 --> 16:09And one of the things that my
16:09 --> 16:11group specialized in was not to
16:11 --> 16:13look at it as just like, oh,
16:13 --> 16:16cancer causing a driver of cancer or a
16:16 --> 16:18passenger that doesn't really cause cancer,
16:18 --> 16:20but quantifying how much each
16:20 --> 16:23mutation is contributing to cancer.
16:23 --> 16:26And the way that sort of came about
16:26 --> 16:27scientifically is many people worked
16:27 --> 16:30on looking at how frequently you
16:30 --> 16:33saw a given mutation in the genome in a
16:33 --> 16:36tumor compared to in normal situations.
16:36 --> 16:40And then what we did is better understand
16:40 --> 16:42the underlying mutational variation
16:42 --> 16:46from site to a site that allows us to
16:46 --> 16:48quantify how much more a certain
16:48 --> 16:50mutation is causing cancer than say another.
16:50 --> 16:52That quantification enables a more
16:52 --> 16:54nuanced view that is not just like,
16:54 --> 16:55oh,
16:55 --> 16:57this is the driver mutation
16:57 --> 16:58causing your cancer.
16:58 --> 16:59And that's the only thing we need
16:59 --> 17:00to know about,
17:00 --> 17:02but rather as I said
17:02 --> 17:05earlier in this discussion,
17:05 --> 17:07what the trajectory of changes
17:07 --> 17:10are and how each one changes your
17:10 --> 17:13prospects going forward with cancer.
17:13 --> 17:15And the key division there is
17:15 --> 17:18between two forces which we ended
17:18 --> 17:20up talking about near the end of
17:20 --> 17:22the our previous talk which is
17:25 --> 17:27there's the underlying mutations that happen.
17:27 --> 17:28What causes those mutations happen
17:28 --> 17:31and on the other hand there's the
17:31 --> 17:33selection or the fact that
17:33 --> 17:34those mutations may increase the
17:34 --> 17:37proliferation or the survival of cancer.
17:37 --> 17:37Of course,
17:37 --> 17:39we don't want cancer to proliferate
17:39 --> 17:40and survive.
17:40 --> 17:42And so the prospect for whether or not,
17:42 --> 17:43say,
17:43 --> 17:45a given drug that you're on is
17:45 --> 17:48under development may or may not
17:48 --> 17:49help a patient if it's targeted
17:49 --> 17:50at a specific driver
17:50 --> 17:52mutation is basically proportional
17:52 --> 17:54to how much it makes that cancer
17:54 --> 17:56cell survivor proliferate better.
17:56 --> 17:59So this quantitative measure that
17:59 --> 18:01we're taking actually tells us the
18:01 --> 18:03prospects for how powerful a
18:03 --> 18:06prospective drug could possibly
18:06 --> 18:09be if it completely abrogates the
18:09 --> 18:11function of the mutated protein.
18:11 --> 18:13And then what we've moved on to
18:13 --> 18:16doing is not just quantifying for
18:16 --> 18:18each individual mutation just what
18:18 --> 18:20the quantitative benefit to the
18:20 --> 18:22cancer cell is or the detriment
18:22 --> 18:25to the patient obviously,
18:25 --> 18:27but quantifying how that benefit
18:27 --> 18:29or detriment changes with other
18:29 --> 18:32mutations that also happening.
18:34 --> 18:36So it's not just a simple change of
18:36 --> 18:38a single gene that leads to cancer.
18:38 --> 18:39In most cases,
18:39 --> 18:41it's usually a cascade of changes.
18:41 --> 18:44And how that cascade plays out determines
18:44 --> 18:47the time course of one's cancer journey.
18:47 --> 18:50And so the more we can better
18:50 --> 18:52understand the genetics underlying that
18:52 --> 18:55journey from a molecular standpoint,
18:55 --> 18:58the better we can understand what the
18:58 --> 19:00patient's journey is going to be and
19:00 --> 19:02treat that patient so that they can
19:02 --> 19:04receive the best outcome possible.
19:05 --> 19:08You know, as you talk about these
19:08 --> 19:12cancers and the mutations and
19:12 --> 19:14how these mutations ultimately
19:14 --> 19:17lead to cancer and how you're able to
19:17 --> 19:19use kind of these mathematical models
19:19 --> 19:23to predict the trajectory.
19:23 --> 19:26I started thinking about cancer in the
19:26 --> 19:29context of the human environment
19:29 --> 19:33in which they are and how different
19:33 --> 19:36that can be in every individual.
19:36 --> 19:39So we know for example that your
19:39 --> 19:43immune system plays a role in terms of
19:43 --> 19:46identifying cells that are thought to
19:46 --> 19:50be quote foreign or mutated including
19:50 --> 19:54cancer cells and how cancer cells
19:54 --> 19:57have started to develop a kind of
19:57 --> 20:00evasion of the immune system.
20:00 --> 20:03And so can you talk a little bit
20:03 --> 20:05about how your mathematical models
20:05 --> 20:09kind of factor in the host in
20:09 --> 20:12terms of the interplay of its
20:12 --> 20:14ability to identify these mutations
20:14 --> 20:17and get rid of them versus not?
20:18 --> 20:20A recent graduate student
20:20 --> 20:22in my laboratory who's now an assistant
20:22 --> 20:25professor at the University of Rhode Island,
20:25 --> 20:27Nick Fisk did some very interesting
20:27 --> 20:29work that is still pre publication.
20:29 --> 20:32But I'm happy to talk about it here where
20:32 --> 20:35we were able to actually look at
20:35 --> 20:37the increase in selection or the amount
20:37 --> 20:40that it benefits cancer or hurts cancer
20:40 --> 20:42to have these particular mutations.
20:42 --> 20:44And we could show a correlation between
20:44 --> 20:47the immune system or the microenvironment,
20:47 --> 20:50how that that microenvironment is responding
20:50 --> 20:52and these selection coefficients themselves.
20:52 --> 20:55So in other words the more the immune
20:55 --> 20:58system could grab on to a particular
20:58 --> 21:01mutation that identifies cancer as
21:01 --> 21:04problematic,
21:04 --> 21:06the more we could see the active
21:06 --> 21:08selection against that particular
21:08 --> 21:10mutation in the individual.
21:10 --> 21:12So at the same time as we're thinking
21:12 --> 21:15about these selection coefficients or
21:15 --> 21:17these benefits or detriments,
21:17 --> 21:19benefits of the cells,
21:19 --> 21:20detriments of the patient,
21:20 --> 21:22of the cancer,
21:22 --> 21:24we can actually look at how that
21:24 --> 21:26interaction is playing into the
21:26 --> 21:28particular mutations that spread or
21:28 --> 21:31don't spread within the cancer cells.
21:31 --> 21:32And that interaction is a really,
21:32 --> 21:35really key thing to understand for many
21:35 --> 21:37different therapies that are being developed
21:39 --> 21:41in immunotherapy areas,
21:41 --> 21:43which is of course a very promising
21:43 --> 21:46area right now in cancer treatment.
21:46 --> 21:48So hopefully what we can do is to
21:48 --> 21:51use those same kinds of measurements
21:51 --> 21:53of how much this allows cells
21:53 --> 21:55to proliferate or survive,
21:55 --> 21:57to better understand which immunotherapies
21:57 --> 22:00are actually going to serve patients
22:00 --> 22:03to a better level as well.
22:03 --> 22:05All of these methods, you know,
22:05 --> 22:08depend on mathematics.
22:08 --> 22:09Of course,
22:09 --> 22:11like any of this sort
22:11 --> 22:12of bioinformatics,
22:12 --> 22:14relies on a lot of algorithms.
22:14 --> 22:16But my collaborator for this most
22:16 --> 22:18recent work looking at the epistasis
22:18 --> 22:19between different mutations,
22:19 --> 22:22Jorge Alfaro Morello,
22:22 --> 22:26is actually a mathematician by training.
22:26 --> 22:27I'm a biologist by training, and
22:27 --> 22:29used a lot of mathematics myself
22:29 --> 22:31in much of my work and early on
22:31 --> 22:33in the development of this most
22:33 --> 22:35recent work I sat down and
22:35 --> 22:37was like OK I really need to look
22:37 --> 22:39not just at individual mutations in
22:39 --> 22:41individual genes as working completely
22:41 --> 22:43independently from everything else
22:43 --> 22:45but as in a pair wise way looking at
22:45 --> 22:48how this gene interacts with this other genes.
22:48 --> 22:50Fortunately I was able to do a little
22:50 --> 22:51bit of mathematics that solved that
22:51 --> 22:53pair wise case and was very proud
22:53 --> 22:54of myself for doing that but
22:54 --> 22:56I ran into a roadblock when I tried
22:56 --> 22:59to look at not just one interaction,
22:59 --> 23:001 gene interacting with another,
23:00 --> 23:01but you know,
23:01 --> 23:03how about those two genes
23:03 --> 23:05interacting with a third gene?
23:05 --> 23:07It starts getting more and more complicated,
23:07 --> 23:08the mathematics that we have to
23:08 --> 23:10use to solve that kind of problem.
23:10 --> 23:13And so I worked with Jorge Alfaro Amarillo,
23:13 --> 23:16who's a research scientist here at Yale,
23:16 --> 23:20and he was able to solve it for
23:20 --> 23:213-4, even 5 different mutations
23:21 --> 23:24and even more given enough data.
23:24 --> 23:26So, we're able to now better
23:26 --> 23:28understand how all of these genes
23:28 --> 23:30are interacting with each other
23:30 --> 23:32during that time course of cancer.
23:32 --> 23:34And that understanding I think
23:34 --> 23:36is going to be critical toward
23:36 --> 23:37the most powerful precision
23:37 --> 23:39medicine we can do in the future.
23:40 --> 23:44So Jeff, you used a term earlier which
23:44 --> 23:46many of us may not be familiar with.
23:46 --> 23:48What exactly is epistasis?
23:49 --> 23:52Yeah you may have remembered
23:52 --> 23:53something from like high school
23:53 --> 23:55genetics or when you learned
23:55 --> 23:57about the peas and the pods and how
23:57 --> 23:59they're different colors and
23:59 --> 24:00stuff and there's something called
24:00 --> 24:03epistasis and what it
24:03 --> 24:05means is just 1 gene is affecting
24:05 --> 24:07what you see in the other genes.
24:07 --> 24:10So it means that you don't necessarily
24:10 --> 24:12get your segregation of three to
24:12 --> 24:15one or 9:00 to 3:00 to 3:00 to 1:00 if
24:15 --> 24:16you remember your high school genetics
24:16 --> 24:18that you expect because there's some
24:18 --> 24:21other gene affecting that segregation.
24:21 --> 24:24So epistasis is just a
24:24 --> 24:25complicated word for
24:25 --> 24:26a fairly simple phenomenon,
24:26 --> 24:27which is just that
24:27 --> 24:31it matters what genetic context you're in.
24:31 --> 24:35Meaning if you have Gene A in a certain form,
24:35 --> 24:37then that's going to change how
24:37 --> 24:39Gene B is going to act or how Gene
24:39 --> 24:41B is going to impact something.
24:41 --> 24:43And in the particular case we're looking at,
24:43 --> 24:46what we're concerned is how much is gene
24:46 --> 24:48A contributing to cancer in general.
24:48 --> 24:51And then the other complication
24:51 --> 24:54that's driven by epistasis is
24:54 --> 24:57what if we have Gene B mutated first,
24:57 --> 24:59how much will A contribute then?
24:59 --> 25:01And in some cases if you have
25:01 --> 25:02Gene B mutated first,
25:02 --> 25:04Gene A won't contribute anything to cancer.
25:05 --> 25:06And in other cases if you
25:06 --> 25:08have Gene B mutated first,
25:08 --> 25:11Gene A contributes much more to cancer.
25:11 --> 25:13And so understanding that is really key.
25:13 --> 25:14So for instance,
25:14 --> 25:17if I were to try to treat
25:17 --> 25:19patients who have gene A mutated,
25:19 --> 25:22depending on which of those cases it was,
25:22 --> 25:24it might really make a big
25:24 --> 25:25difference to whether that therapy
25:25 --> 25:27might actually be beneficial.
25:27 --> 25:31And so this is actually a tool in part for
25:31 --> 25:33identifying biomarkers that mean
25:33 --> 25:35therapy towards this gene might work,
25:35 --> 25:37but we need to know about this other
25:37 --> 25:38gene to know whether it'll work.
25:39 --> 25:42It kind of almost makes
25:42 --> 25:45me think that if you could identify
25:45 --> 25:48that mutations in gene A cause cancer.
25:48 --> 25:51But if you have mutation in Gene B,
25:51 --> 25:55then mutations in gene A will not lead to
25:55 --> 25:58the development of a full blown cancer.
25:58 --> 26:00That you could potentially develop
26:00 --> 26:03a screening tool for patients
26:03 --> 26:05who have gene A mutations.
26:05 --> 26:06And in those patients,
26:06 --> 26:10you might be able to create a cellular
26:10 --> 26:13therapy where you induce mutation in gene B,
26:13 --> 26:16which then turns off the
26:16 --> 26:18effect of mutations in Gene A.
26:18 --> 26:21Is that kind of where you're going with this?
26:21 --> 26:23Certainly with enough data we
26:23 --> 26:24can get at questions like that.
26:24 --> 26:26Right now, it's a little hard
26:26 --> 26:28for us to understand the sort of
26:28 --> 26:29negative interactions very well.
26:29 --> 26:31We mostly understand the positive
26:31 --> 26:32interactions
26:32 --> 26:34but I think as we get more and
26:34 --> 26:36more data and it is
26:36 --> 26:38churning out even every six months,
26:38 --> 26:40I sort of look back at how much data
26:40 --> 26:42we have on each different cancer
26:42 --> 26:44and the amount it's increasing
26:44 --> 26:46is just astounding and wonderful
26:46 --> 26:48for our kind of science.
26:48 --> 26:49So I think in time we're
26:49 --> 26:50going to be able to get at
26:50 --> 26:52questions like that where we'll
26:52 --> 26:54be able to say look, if you get
26:54 --> 26:57rid of the function of this gene,
26:57 --> 26:58then this other gene won't have
26:58 --> 27:00the impact that it would otherwise.
27:00 --> 27:01And that may be a really,
27:01 --> 27:02really,
27:02 --> 27:04really beneficial way to sort
27:04 --> 27:06of guide our therapeutic
27:06 --> 27:07discovery.
27:08 --> 27:11So thinking about the future,
27:11 --> 27:13what things are you working on now
27:13 --> 27:15and what things are you really
27:15 --> 27:16excited about in terms of where
27:16 --> 27:18this field is going in the future?
27:19 --> 27:21As is typical in
27:21 --> 27:23science in my group, we tend to
27:23 --> 27:26use the tools that we have
27:26 --> 27:27available which are somewhat unique,
27:27 --> 27:30but to address the things that can be
27:30 --> 27:32addressed before the things that are much,
27:32 --> 27:33much harder to address.
27:33 --> 27:35And what we focused on mostly
27:35 --> 27:37are these individual changes in
27:37 --> 27:39individual base pairs of the DNA
27:39 --> 27:40that lead to a change in an amino
27:40 --> 27:43acid and then cause proteins to
27:43 --> 27:47function in ways that lead to cancer.
27:47 --> 27:49But there's a a whole suite of other
27:49 --> 27:51kinds of changes that occur that are
27:51 --> 27:53well known to be important to cancer.
27:53 --> 27:55So for instance,
27:55 --> 27:57in addition to the typical,
27:57 --> 27:59you know, base pair change in DNA
27:59 --> 28:01that leads to amino acid changes,
28:01 --> 28:04you can have something called methylation,
28:04 --> 28:06which it means those base pairs get
28:06 --> 28:08sort of tagged with this methyl group
28:08 --> 28:10and it means that the those genes that
28:10 --> 28:12have that methylation are either not
28:12 --> 28:14expressed or in some cases are expressed.
28:14 --> 28:16It depends on exactly the context.
28:16 --> 28:18But that methylation process is
28:18 --> 28:20known to be relevant to cancer,
28:20 --> 28:21and so understanding how those
28:21 --> 28:23contribute to cell proliferation and
28:23 --> 28:25survival in the same depth that we
28:25 --> 28:26understand these single nucleotide
28:26 --> 28:28mutations is a major goal in our group.
28:29 --> 28:31Doctor Jeffrey Townsend is the Eliu
28:31 --> 28:34Professor of Biostatistics and professor
28:34 --> 28:36of Ecology and Evolutionary biology
28:36 --> 28:38at the Yale School of Medicine.
28:38 --> 28:40If you have questions,
28:40 --> 28:42the address is canceranswers@yale.edu,
28:42 --> 28:45and past editions of the program
28:45 --> 28:47are available in audio and written
28:47 --> 28:48form at yalecancercenter.org.
28:48 --> 28:51We hope you'll join us next week to
28:51 --> 28:53learn more about the fight against
28:53 --> 28:55cancer here on Connecticut Public Radio.
28:55 --> 28:57Funding for Yale Cancer Answers is
28:57 --> 29:00provided by Smilow Cancer Hospital.

Download Audio Download Transcript

Information

Tracking Cancer Cells with guest Dr. Jeffrey Townsend January 28, 2024

11231

Guests

Dr. Jeffrey Townsend

To Cite

DCA Citation Guide