Classical Hematology

Transcript

00:00 --> 00:03Funding for Yale Cancer Answers is
00:03 --> 00:06provided by Smilow Cancer Hospital.
00:06 --> 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 --> 00:14by welcoming oncologists and
00:14 --> 00:17specialists who are on the forefront
00:17 --> 00:18of the battle to fight cancer.
00:18 --> 00:20This week it's a conversation about
00:20 --> 00:22the field of classical hematology
00:22 --> 00:24with Doctor George Goshua.
00:24 --> 00:26Dr Goshua is an assistant professor
00:26 --> 00:28of medicine and hematology at
00:28 --> 00:29the Yale School of Medicine,
00:29 --> 00:31where Doctor Chagpar is a
00:31 --> 00:33professor of surgical oncology.
00:34 --> 00:35George, maybe we can start off by
00:35 --> 00:37you telling us a little bit more
00:37 --> 00:39about yourself and what it is you do.
00:39 --> 00:41Of course, it would be my pleasure.
00:41 --> 00:44I am a classical hematologist
00:44 --> 00:46by training and methodologically
00:46 --> 00:48I'm trained in decision science,
00:48 --> 00:50so I'm also a decision scientist.
00:50 --> 00:53And on faculty here at Yale
00:53 --> 00:55University School of Medicine,
00:55 --> 00:56I run the Goshua lab,
00:56 --> 00:59which is a quantitative decision
00:59 --> 01:00analytic modeling lab,
01:01 --> 01:03the first in the country to
01:03 --> 01:05focus on classical hematology.
01:05 --> 01:07And I have the privilege of working
01:07 --> 01:08with undergraduate students,
01:08 --> 01:10graduate students at the School
01:10 --> 01:11of Public Health,
01:11 --> 01:13the School of Medicine,
01:13 --> 01:14and beyond.
01:14 --> 01:18Many of us have heard about hematology,
01:18 --> 01:22but what exactly is classical hematology?
01:22 --> 01:24It seems to remind me about classical
01:24 --> 01:27music as opposed to music in general.
01:27 --> 01:30So tell us more about what exactly is
01:30 --> 01:33classical hematology and how that varies
01:33 --> 01:35from all other forms of hematology.
01:36 --> 01:39I'm really glad you asked that question.
01:39 --> 01:42And that's because the field has really
01:42 --> 01:45struggled with its name until very recently.
01:45 --> 01:48The American Society of Hematology
01:48 --> 01:52has put forward a campaign to
01:52 --> 01:55unify the field and call
01:55 --> 01:57it classical hematology.
01:57 --> 01:59And the way that it differs from other
01:59 --> 02:02hematology is that we take care of patients
02:02 --> 02:04with non cancerous blood disorders.
02:04 --> 02:06And the reason why the naming matters in
02:06 --> 02:09particular is the other names for the field.
02:09 --> 02:11There's two. There is non
02:11 --> 02:11malignant hematology,
02:11 --> 02:15so non cancerous and then benign
02:15 --> 02:18hematology which is quite common and
02:18 --> 02:20that latter term is particularly problematic
02:23 --> 02:27because as we probably will discuss here,
02:27 --> 02:30a lot of our patients have
02:30 --> 02:32life altering diseases that they
02:32 --> 02:35have to live with and in some cases
02:35 --> 02:38very deadly diseases that can be
02:38 --> 02:39deadly without appropriate treatment.
02:39 --> 02:41And so for that reason there has been
02:41 --> 02:44also a lot of frustration on our
02:44 --> 02:46patients part with regards to being
02:46 --> 02:48labeled as quote unquote benign.
02:48 --> 02:50And so for that reason the field has
02:50 --> 02:52moved forward now just this year
02:52 --> 02:53actually with classical hematology.
02:54 --> 02:56So give us some examples of
02:56 --> 03:00some of the not malignant
03:00 --> 03:02hematologic disorders that you treat.
03:03 --> 03:05Of course, there's a lot
03:05 --> 03:06of rare diseases in here,
03:06 --> 03:09but there's also less rare diseases too.
03:09 --> 03:11And so maybe I'll start with diseases
03:11 --> 03:13that folks might be more familiar with,
03:13 --> 03:15even though some of them are still rare.
03:15 --> 03:17Sickle cell disease in particular, right?
03:17 --> 03:19I think a lot of us know individuals
03:19 --> 03:21who live with sickle cell disease.
03:21 --> 03:23But then as we move forward,
03:23 --> 03:26think about all of your
03:26 --> 03:28auto immune conditions,
03:28 --> 03:30so conditions where the immune
03:30 --> 03:31system is dysregulated.
03:31 --> 03:33And then that causes derangements
03:33 --> 03:34in the blood parameters.
03:34 --> 03:38And so these are diseases in the realm of one
03:38 --> 03:41to three in a million in terms of incidence.
03:41 --> 03:43And examples include paroxysmal
03:43 --> 03:45nocturnal hemoglobinuria,
03:45 --> 03:48immune thrombotic thrombocytopenic purpura,
03:48 --> 03:50chronic immune thrombocytopenia,
03:50 --> 03:51porphyrias.
03:51 --> 03:53And then when you think
03:53 --> 03:54about more common things,
03:54 --> 03:56venous thromboembolism which affects hundreds
03:56 --> 03:59of thousands of Americans every year,
03:59 --> 04:01iron deficiency anemia
04:01 --> 04:04which affects a lot of our men and women,
04:04 --> 04:06and in particular pregnant
04:06 --> 04:08women as well in this country.
04:08 --> 04:10So that's a little bit
04:10 --> 04:12of a sampling of the more rare
04:12 --> 04:13and then the more common.
04:15 --> 04:18It really seems to be a
04:18 --> 04:21wide spectrum of disease.
04:21 --> 04:23And is the only linkage
04:23 --> 04:25between all of them that they
04:25 --> 04:27have to have something to do
04:27 --> 04:29with blood and blood disorders?
04:29 --> 04:31I think that's very fair to say.
04:31 --> 04:34Yeah, it's interesting because at least in
04:34 --> 04:39the case of let's say autoimmune disorders.
04:39 --> 04:41Sometimes in some of them if
04:41 --> 04:43you want to think about it,
04:43 --> 04:45this is how I think about it with my
04:45 --> 04:47patients when we talk together
04:47 --> 04:49in clinic, you can think about it as
04:49 --> 04:51the disease spilling over into the
04:51 --> 04:54blood and the blood is very sensitive.
04:54 --> 04:56We have multiple cell lines
04:56 --> 04:57that can be affected.
04:57 --> 05:00We have multiple proteins floating
05:00 --> 05:03in there and our immune system
05:03 --> 05:05that has been so finely tuned over
05:05 --> 05:07over millennia and any of these
05:07 --> 05:09parameters can be thrown off.
05:09 --> 05:11And so I think it's very fair to
05:11 --> 05:13say that the commonality here is
05:13 --> 05:15that there's some underlying issue
05:15 --> 05:17that's happening to one of or more
05:17 --> 05:18of those parameters in the blood.
05:19 --> 05:22Because it also seems that
05:22 --> 05:25when you're thinking about things as
05:25 --> 05:30diverse as ITP versus sickle cell,
05:30 --> 05:33anemia versus thromboembolism,
05:33 --> 05:36the treatments are very different.
05:36 --> 05:39The patient populations are very different.
05:39 --> 05:41Even the blood cells that are
05:41 --> 05:42affected are very different.
05:44 --> 05:45That's exactly correct.
05:45 --> 05:49There's a beautiful diversity and
05:49 --> 05:51heterogeneity within the field.
05:51 --> 05:54There are classical hematologists who
05:54 --> 05:56particularly focus or sub-specialize
05:56 --> 05:58further even within that field.
05:58 --> 06:00That is part of the reason why,
06:00 --> 06:01because there is such a diversity.
06:01 --> 06:04And then there are other classical
06:04 --> 06:06hematologists who are more generalists
06:06 --> 06:08as they would be in any specialty
06:08 --> 06:10that kind of see the full spectrum
06:10 --> 06:12and then if there are complications
06:12 --> 06:14or there's a particularly
06:14 --> 06:16high risk situation,
06:16 --> 06:17in those circumstances,
06:17 --> 06:21they will often refer to a tertiary
06:21 --> 06:24academic Center for further evaluation.
06:24 --> 06:28George, many of us may
06:28 --> 06:31be familiar with some of these
06:31 --> 06:33blood disorders that you mentioned,
06:33 --> 06:35but you also mentioned that you
06:35 --> 06:38have a laboratory that focuses on
06:38 --> 06:41quantitative modeling and decision
06:41 --> 06:44analytics that seems to be very
06:44 --> 06:46different from what we would
06:46 --> 06:48normally think of as a hematologist.
06:48 --> 06:52Tell us more about how those two
06:52 --> 06:54areas of interest and expertise
06:54 --> 06:56kind of merged for you.
06:58 --> 07:00Well, I think it has a lot to
07:00 --> 07:02do with advocacy. By definition,
07:02 --> 07:06a lot of our diseases are rare in
07:06 --> 07:08our field all across the spectrum.
07:08 --> 07:10When you combine them all together,
07:10 --> 07:12you really get very
07:12 --> 07:14significant numbers of individuals.
07:14 --> 07:15But within each bin,
07:15 --> 07:18if we want to think about it that way,
07:18 --> 07:21some of the diseases are particularly rare.
07:21 --> 07:23And it is for that reason that you start
07:23 --> 07:25to think more and more about decision
07:25 --> 07:28making in an area where there are a lot of
07:28 --> 07:31diseases that are rare and in an
07:31 --> 07:34area where there are, let's say,
07:34 --> 07:37less prospective randomized clinical trials,
07:37 --> 07:41perhaps more of a dependence
07:41 --> 07:43on observational data,
07:43 --> 07:46you start to think about trying to make
07:46 --> 07:49decisions with your patients in the
07:49 --> 07:52clinic and in the hospital in some cases,
07:52 --> 07:55some of which have very significant
07:55 --> 07:58consequences or can have very significant
07:58 --> 08:00consequences on the rest of their lives.
08:00 --> 08:03We use strong immunosuppressive agents.
08:03 --> 08:06We use anticoagulation,
08:06 --> 08:09blood thinners that can predispose
08:09 --> 08:11people if using correctly,
08:11 --> 08:16unnecessarily to a risk of bleeding and so
08:16 --> 08:20it feels very natural to try and
08:20 --> 08:23quantitatively try to approach
08:23 --> 08:27these decisions and put them in a
08:27 --> 08:30framework that matters to patients,
08:30 --> 08:31to physicians,
08:31 --> 08:34to payers and then try to push
08:34 --> 08:37the care of patients forward.
08:37 --> 08:40And decision science is really nice
08:40 --> 08:42because one of the very wonderful
08:42 --> 08:44and unique things about it is
08:44 --> 08:46it's very explicit in its
08:46 --> 08:49measurement and reporting of uncertainty and
08:49 --> 08:53so any decision that we make in our lives,
08:53 --> 08:55anytime you think of a trade off
08:55 --> 08:56and I think about trade-offs all
08:56 --> 08:59of the time, decision scientists do,
08:59 --> 09:01but everyone does beyond decision
09:01 --> 09:01scientists too,
09:02 --> 09:02right?
09:02 --> 09:04It doesn't have to apply to medicine
09:04 --> 09:06every time you think of a trade off.
09:06 --> 09:10And the downstream effects thereof,
09:10 --> 09:12all of that can be captured and that's
09:12 --> 09:14the really exciting part because I
09:14 --> 09:16think we have an opportunity to move
09:16 --> 09:18the care of these patients forward and help improve
09:19 --> 09:22the areas of our health system,
09:22 --> 09:24and there are many that need improvement.
09:26 --> 09:28And so it sounds like you know this
09:28 --> 09:30whole area of decision science
09:30 --> 09:32would have broad applicability
09:32 --> 09:35to all fields of medicine really
09:35 --> 09:38where we're balancing as you say
09:38 --> 09:40trade-offs between risks and benefits
09:40 --> 09:43and how each patient might value
09:43 --> 09:46each of those things differently.
09:46 --> 09:49Talk a little bit more about kind of
09:49 --> 09:52the practical examples of how you
09:52 --> 09:55applied decision science in your clinical
09:55 --> 09:56endeavors.
09:58 --> 10:02Of course. We'll start with an
10:02 --> 10:05earlier example
10:05 --> 10:08and then I'll work my way forward.
10:08 --> 10:10So anytime you think of a decision problem,
10:10 --> 10:12and you think of trade-offs,
10:12 --> 10:14you want to be able to make sure that you
10:14 --> 10:17have it laid out clearly in front of you.
10:17 --> 10:20And so I'm going to use a very
10:20 --> 10:22interesting problem because it employs
10:22 --> 10:243 different strategies in a disease
10:24 --> 10:27where your platelet counts are low,
10:27 --> 10:28chronic immune thrombocytopenia.
10:28 --> 10:31When your platelet counts are low,
10:31 --> 10:33you're at an increased risk of bleeding.
10:33 --> 10:36And for that reason
10:36 --> 10:38there are treatment options and
10:38 --> 10:40treatments that we do pursue for
10:40 --> 10:42individuals whose platelet
10:42 --> 10:44counts are particularly low because
10:44 --> 10:47we don't want them to have a bleed,
10:47 --> 10:49especially if it's a bleed in the head,
10:49 --> 10:51sometimes a bleed in the gut,
10:52 --> 10:54the bleeding can really happen anywhere,
10:54 --> 10:56but there are certain higher risk areas.
10:56 --> 10:59And so in thinking through that,
10:59 --> 11:01by the time an individual has,
11:01 --> 11:02let's say,
11:02 --> 11:04a diagnosis of immune thrombocytopenia,
11:04 --> 11:05by the time they reach 12 months,
11:05 --> 11:07it's defined as chronic.
11:07 --> 11:09It's done that way because there's a
11:09 --> 11:11subset of individuals who will never go on
11:11 --> 11:13to develop chronic immune thrombocytopenia.
11:13 --> 11:15Their platelet counts will improve,
11:15 --> 11:16sometimes even spontaneously and
11:16 --> 11:19sometimes with a little bit of treatment,
11:19 --> 11:22and they will no longer need any treatment.
11:22 --> 11:23But for the vast majority of
11:23 --> 11:26individuals who do get to the stage
11:26 --> 11:28of having one year of this disease,
11:28 --> 11:31now they have a chronic disease and within
11:31 --> 11:34we know the Natural History of that disease
11:35 --> 11:35at that point,
11:35 --> 11:37it's much less likely that
11:37 --> 11:38it's going to dissipate.
11:38 --> 11:40And so often these
11:40 --> 11:41individuals need treatment.
11:41 --> 11:43And so the treatment decision
11:43 --> 11:43here is fascinating.
11:43 --> 11:46And this is 1 classic example where
11:46 --> 11:48a randomized control trial will
11:48 --> 11:50never be done for reasons that
11:50 --> 11:51will become clear in a moment.
11:51 --> 11:55And that is the fact that our treatment
11:55 --> 11:57options include three options here.
11:57 --> 11:59And they include a surgical approach,
11:59 --> 12:02splenectomy to try and remove the
12:02 --> 12:04spleen and remove a site of production.
12:04 --> 12:07Of all of these auto antibodies that
12:07 --> 12:10are in part driving the disease process.
12:10 --> 12:13And we know that about 60% of
12:13 --> 12:15individuals will then never have to
12:15 --> 12:19think or worry about this disease again.
12:19 --> 12:20At the same time,
12:20 --> 12:22splenectomy carries the risks
12:22 --> 12:24of infection that are lifelong.
12:24 --> 12:25Although they are time variant,
12:25 --> 12:27they change over time.
12:27 --> 12:30It carries a risk of developing a
12:30 --> 12:32blood clot overtime going forward
12:32 --> 12:34and that's also time variant that
12:34 --> 12:35changes with time.
12:35 --> 12:38And separately anytime you perform surgery
12:38 --> 12:41there is a risk of having complications.
12:41 --> 12:44And even deaths from the surgery itself.
12:44 --> 12:47And so you think about a strategy
12:47 --> 12:49like that versus thinking about the
12:49 --> 12:51two other options which include
12:51 --> 12:52thrombopoietin receptor agonists,
12:52 --> 12:54which are these therapies
12:54 --> 12:56that are taken chronically,
12:56 --> 12:58either intravenously or by
12:58 --> 13:02mouth as tablets and
13:02 --> 13:04technically have been studied going
13:04 --> 13:06forward and thinking about using
13:06 --> 13:08them for a prolonged period of time,
13:08 --> 13:11so not just a few weeks or a few
13:11 --> 13:13months with the idea being that
13:13 --> 13:14you might have to be on
13:14 --> 13:15this therapy lifelong.
13:15 --> 13:17There are certain very
13:17 --> 13:18expensive costs of course,
13:18 --> 13:19that accrue with this therapy,
13:19 --> 13:23both to the health system and to patients.
13:23 --> 13:25And about 1/3 of patients at a median of
13:25 --> 13:282 1/2 years can come off of therapy and
13:28 --> 13:30probably be successful
13:30 --> 13:32though we don't have enough follow
13:32 --> 13:34up time to know for sure and then
13:34 --> 13:36separate from that in the last third
13:38 --> 13:40is an immunosuppressive agent called
13:40 --> 13:42Rituximab that depletes those cells
13:42 --> 13:44that produce those troublesome auto
13:44 --> 13:46antibodies and you have response
13:47 --> 13:48in about 50% of individuals
13:48 --> 13:51at about a year.
13:51 --> 13:53And then that response starts to degrade,
13:53 --> 13:55it starts to decrease,
13:55 --> 13:57and people will have relapses.
13:57 --> 13:58And so if you can imagine,
13:58 --> 14:00you have these three options.
14:00 --> 14:01But in truth,
14:01 --> 14:03you can also sequence these options.
14:03 --> 14:06And if you look at the American
14:06 --> 14:08Society of Hematology guidelines,
14:08 --> 14:10there's this inherent struggle with
14:10 --> 14:12how do you actually rank these options
14:12 --> 14:14when they have not been compared
14:14 --> 14:17head-to-head and who is going to be
14:17 --> 14:18randomizing people to receive surgery,
14:18 --> 14:19splenectomy versus not,
14:19 --> 14:22that's not going to happen.
14:22 --> 14:24But we do have 20 years of follow-up data
14:24 --> 14:28with this modality with surgery specifically.
14:28 --> 14:29And in the clinics
14:29 --> 14:32we can see that over the last 20 years,
14:32 --> 14:35the utilization of surgery has significantly
14:35 --> 14:38gone down in part because of these newer,
14:38 --> 14:40more expensive therapies,
14:40 --> 14:41not because
14:41 --> 14:44a splenectomy is not an effective option.
14:44 --> 14:49And so that is a perfect setup then and
14:49 --> 14:51framework to start thinking about how
14:51 --> 14:54do we actually accurately model this,
14:54 --> 14:55how do we show what the benefit is
14:55 --> 14:57on a population level and then can
14:57 --> 14:59we also make it covariate specific?
14:59 --> 15:01Meaning if you look at the
15:01 --> 15:02specific comorbidities,
15:02 --> 15:04IE the diseases that the patients
15:04 --> 15:05have and their likeliness to
15:05 --> 15:07respond to one of these therapies,
15:07 --> 15:09can we build that in further than
15:09 --> 15:11to try and make it an individualized
15:11 --> 15:13personalized treatment decision for them?
15:14 --> 15:16We'll pick up that conversation,
15:16 --> 15:18but first we need to take a
15:18 --> 15:20short break for a medical minute.
15:20 --> 15:23Please stay tuned to learn more about
15:23 --> 15:25classical hematology with my guest,
15:25 --> 15:26Doctor George Goshua.
15:26 --> 15:28Funding for Yale Cancer Answers
15:28 --> 15:31comes from Smilow Cancer Hospital,
15:31 --> 15:33where their Center for Gastrointestinal
15:33 --> 15:34Cancers provides patients with
15:34 --> 15:36gastric cancers a comprehensive,
15:36 --> 15:37multidisciplinary approach to
15:37 --> 15:39the treatment of their cancer,
15:39 --> 15:42including clinical trials.
15:42 --> 15:45Smilowcancerhospital.org.
15:45 --> 15:48Over 230,000 Americans will be
15:48 --> 15:50diagnosed with lung cancer this year,
15:50 --> 15:52and in Connecticut alone there
15:52 --> 15:55will be over 2700 new cases.
15:55 --> 15:57More than 85% of lung cancer
15:57 --> 15:59diagnosis are related to smoking,
15:59 --> 16:02and quitting even after decades of use,
16:02 --> 16:04can significantly reduce your risk
16:04 --> 16:06of developing lung cancer each day.
16:06 --> 16:09Patients with lung cancer are surviving
16:09 --> 16:11thanks to increased access to advanced
16:11 --> 16:13therapies and specialized care.
16:13 --> 16:14New treatment options and
16:14 --> 16:16surgical techniques are giving
16:16 --> 16:17lung cancer survivors more hope
16:17 --> 16:19than they have ever had before.
16:19 --> 16:22Clinical trials are currently underway
16:22 --> 16:24at federally designated Comprehensive
16:24 --> 16:26cancer centers such as the battle
16:26 --> 16:28two trial at Yale Cancer Center and
16:28 --> 16:30Smilow Cancer Hospital to learn if a
16:30 --> 16:33drug or combination of drugs based
16:33 --> 16:35on personal biomarkers can help to
16:35 --> 16:37control non small cell lung cancer.
16:37 --> 16:40More information is available
16:40 --> 16:41at yalecancercenter.org.
16:41 --> 16:43You're listening to Connecticut public radio.
16:44 --> 16:46Welcome back to Yale Cancer Answers.
16:46 --> 16:48This is doctor Anees Chagpar
16:48 --> 16:50and I'm joined tonight by my guest,
16:50 --> 16:51Doctor George Goshua.
16:51 --> 16:54We're talking about the field of classical
16:54 --> 16:56hematology and more specifically,
16:56 --> 16:59Doctor Goshua has a special
16:59 --> 17:01expertise in decision science.
17:01 --> 17:03And right before the break,
17:03 --> 17:06he was starting to tell us about how he
17:06 --> 17:08brings decision science into the clinic.
17:08 --> 17:10So George, maybe you can pick up
17:10 --> 17:12the conversation where we left it.
17:12 --> 17:14So as I understand
17:14 --> 17:18we were talking about ITP and how there
17:18 --> 17:22are three different options for treatment,
17:22 --> 17:24surgical versus non surgical
17:24 --> 17:27and these can be sequenced.
17:27 --> 17:30We really don't have a lot
17:30 --> 17:31of clinical trial data,
17:31 --> 17:34but you were about to tell us kind
17:34 --> 17:37of how you use decision analytics
17:37 --> 17:40as we come back to this decision of
17:40 --> 17:43splenectomy versus the medication options.
17:43 --> 17:45We know what the data
17:45 --> 17:48looks like at least observationally
17:48 --> 17:50for splenectomy, right.
17:50 --> 17:51We know it's risk profile.
17:51 --> 17:54We know that over the last 20 years
17:54 --> 17:56we've kind of moved away from it and
17:56 --> 17:59I think in some ways for good reason.
17:59 --> 18:01But the question then becomes
18:01 --> 18:04what is that good reason,
18:04 --> 18:06the good reason being that it's often
18:06 --> 18:09assumed I think by us as physicians
18:09 --> 18:11that our patients prefer therapies and
18:11 --> 18:13therapeutics that are less invasive.
18:13 --> 18:15And more often than not,
18:15 --> 18:17that is correct.
18:17 --> 18:20But sometimes there are circumstances
18:20 --> 18:22where patients,
18:22 --> 18:23their values and preferences
18:23 --> 18:25of course are paramount.
18:25 --> 18:26And so sometimes there are
18:26 --> 18:28circumstances where you actually
18:28 --> 18:30will have an individual who is
18:30 --> 18:31interested in pursuing splenectomy.
18:31 --> 18:33In this particular context,
18:33 --> 18:35but will not because of the
18:35 --> 18:37counseling that they receive.
18:37 --> 18:39And so we wanted to take a very
18:39 --> 18:42objective look at this and to model
18:42 --> 18:44what would your life look like,
18:44 --> 18:45you know,
18:45 --> 18:47if you can simulate a thousands of
18:47 --> 18:49times making one decision or another
18:49 --> 18:51decision or yet another decision.
18:51 --> 18:53And that is the beauty of
18:53 --> 18:54decision analytic modeling.
18:54 --> 18:56It allows us to quantify that.
18:56 --> 18:59It allows us to run those simulations
18:59 --> 19:02to make sure that we have addressed
19:02 --> 19:04all of the concerns and so
19:04 --> 19:05putting that all together,
19:05 --> 19:08what we showed was that
19:08 --> 19:09utilizing splenectomy early is
19:09 --> 19:12absolutely fine and in fact the
19:12 --> 19:14quality adjusted life years that you
19:14 --> 19:18accrue if you as the patient make a
19:18 --> 19:20decision to pursue splenectomy at
19:20 --> 19:22least on a population level that
19:22 --> 19:25is just as fine of a decision as
19:25 --> 19:27pursuing the medication therapies.
19:27 --> 19:29And so for those individuals
19:29 --> 19:31for whom it makes sense,
19:31 --> 19:32they shouldn't be dissuaded for
19:32 --> 19:34pursuing a therapy that is going
19:34 --> 19:36to be just as effective for them,
19:36 --> 19:39if the two options are equivalent,
19:39 --> 19:41patients may still be left
19:41 --> 19:43in this decisional conundrum.
19:43 --> 19:46And so how do you help patients with that?
19:46 --> 19:49That drives back to one
19:49 --> 19:52approach that my lab takes is to make
19:52 --> 19:54sure that whenever we build models
19:54 --> 19:56that try to approximate real life
19:56 --> 19:58and that's what they are, right.
19:58 --> 20:00There are only approximations.
20:00 --> 20:02We always take the most
20:02 --> 20:03conservative assumptions.
20:03 --> 20:05And so for example,
20:05 --> 20:06in that particular study,
20:06 --> 20:08although we show equivalence where
20:08 --> 20:11in the past the thought has been
20:11 --> 20:14or the clinical practice has been
20:14 --> 20:16to pursue the medication therapy.
20:16 --> 20:18Although we show equivalence,
20:18 --> 20:21in fact if you use assumptions
20:21 --> 20:23that are more realistic,
20:23 --> 20:25i.e do not downplay the benefits
20:25 --> 20:28of splenectomy and do not over
20:28 --> 20:31exaggerate the risks, which is what
20:31 --> 20:32we did in this model,
20:32 --> 20:35then you'll find that the splenectomy
20:35 --> 20:38option becomes a little bit more
20:38 --> 20:41favorable in certain circumstances.
20:41 --> 20:43But separate from that because we're
20:43 --> 20:45talking on a population level and
20:45 --> 20:47the really exciting bit is that
20:47 --> 20:49we can take that and then we
20:49 --> 20:50can personalize it, right?
20:50 --> 20:52Because this is on a population level,
20:52 --> 20:54this is all comers.
20:54 --> 20:56If you're a 30 year old woman
20:56 --> 20:59versus if you're a 55 year old man,
20:59 --> 21:01there's a very real difference
21:01 --> 21:02in your actual responses,
21:02 --> 21:05a 30 year old woman will have
21:05 --> 21:07a much better outcome,
21:07 --> 21:08typically with splenectomy than
21:08 --> 21:11a 55 year old man as compared
21:11 --> 21:12to the medication therapies.
21:12 --> 21:15And so the next steps for
21:15 --> 21:17that particular question are
21:17 --> 21:19to personalize and
21:19 --> 21:20not just to see,
21:20 --> 21:22but to actually give an opportunity
21:22 --> 21:24for physicians right through
21:24 --> 21:25an easy visual interface,
21:25 --> 21:27essentially where they can plug
21:27 --> 21:29in the parameters of importance
21:29 --> 21:31like age and gender and other
21:31 --> 21:33diseases that may be at play that
21:33 --> 21:34we know affect these risks.
21:34 --> 21:36To then in their clinic calculate
21:36 --> 21:38and simulate what actually
21:38 --> 21:40happened the vast majority of the
21:40 --> 21:42time and to be able to provide
21:42 --> 21:44those estimates to patients so
21:44 --> 21:45they can make a decision that
21:45 --> 21:46makes the most sense for them.
21:48 --> 21:51And that sounds,
21:51 --> 21:54you know, really quite wonderful if
21:54 --> 21:57you're able to take all of the data,
21:57 --> 22:00put it into an analytic model that can
22:00 --> 22:02be personalized so that people can say,
22:02 --> 22:05OK, tell me what's best for me and you
22:05 --> 22:08can put in all of those parameters.
22:08 --> 22:10That sounds really quite wonderful.
22:10 --> 22:15Has that found its way into the clinic in
22:15 --> 22:19hematology specifically, but then if it
22:19 --> 22:22has, where are we going in terms of taking
22:22 --> 22:25that into the clinic for many, many,
22:25 --> 22:28many other diseases where patients still
22:28 --> 22:31struggle with well, what should I do?
22:31 --> 22:33Should I, if I have breast cancer,
22:33 --> 22:35should I have a lumpectomy?
22:35 --> 22:36Should I have a mastectomy,
22:36 --> 22:38should I do one side,
22:38 --> 22:39should I do both sides?
22:39 --> 22:43I mean I can see where this kind of
22:43 --> 22:46modeling would be helpful across diseases.
22:48 --> 22:51Yes. And it has been utilized
22:51 --> 22:53in other disease areas not
22:53 --> 22:55yet in classical hematology,
22:55 --> 22:57but I'm really glad you brought
22:57 --> 22:59up the example of breast cancer.
22:59 --> 23:00The United States Preventative
23:00 --> 23:02Services Task Force,
23:02 --> 23:03their recommendation is actually
23:03 --> 23:05based on micro simulation modeling,
23:05 --> 23:08which is a different kind of
23:08 --> 23:09decision analytic modeling for
23:09 --> 23:11patients with breast cancer.
23:11 --> 23:13Micro simulations have also
23:13 --> 23:15been employed to inform the care
23:15 --> 23:17of patients with lung cancer
23:17 --> 23:18and lung cancer screening.
23:18 --> 23:20So there's a very real opportunity
23:20 --> 23:23here to be able to apply to a
23:23 --> 23:25field where we have diseases that
23:25 --> 23:28are also rare and also quite
23:28 --> 23:30consequential for our patients.
23:30 --> 23:32And that's the exciting part of it too.
23:32 --> 23:34And the exciting bit specifically
23:34 --> 23:36is the fact that
23:36 --> 23:38the decision science methodologists
23:40 --> 23:43have been pushing that field forward for
23:43 --> 23:46many decades now and the opportunity to
23:46 --> 23:49then take the clinical knowledge that
23:49 --> 23:51that we've accumulated as physicians
23:51 --> 23:54and to be able to try and fuse those
23:54 --> 23:56areas of expertise that is what drove
23:56 --> 23:59me to this point because it gives me
23:59 --> 24:01a unique opportunity to work with some
24:01 --> 24:03of the brightest minds and decision
24:03 --> 24:05science and some of the brightest
24:05 --> 24:08minds in clinical medicine to try and
24:08 --> 24:09conceptualize these problems and
24:09 --> 24:11capture them in a way that actually can
24:11 --> 24:14inform one health policy and then second,
24:14 --> 24:16individualized treatment decisions
24:16 --> 24:17for patients.
24:18 --> 24:20So a couple of questions on that.
24:20 --> 24:22So the first question is why hasn't
24:22 --> 24:25it found its way into clinical
24:25 --> 24:27practice in clinical hematology?
24:27 --> 24:29I mean, at the outset you made a
24:29 --> 24:32very nice case for using decision
24:32 --> 24:34science in classical hematology,
24:34 --> 24:37that being that we don't have large
24:37 --> 24:39randomized control trials for what are,
24:39 --> 24:42you know, often rare diseases,
24:42 --> 24:45one would think that this would be an ideal
24:45 --> 24:47platform for the classical hematology.
24:47 --> 24:48So why hasn't it
24:48 --> 24:50found its way into clinical practice yet?
24:51 --> 24:55I think 2 reasons, probably one
24:55 --> 24:58decision science methodologically is,
24:58 --> 25:00I've been told a few times, one of
25:00 --> 25:03the most niche, if not the most niche,
25:03 --> 25:08area speaking methodologically,
25:08 --> 25:11there's just not a lot of decision
25:11 --> 25:13scientists in this country.
25:13 --> 25:14There's a little bit of a hub on the
25:14 --> 25:16West Coast, a little bit in the Midwest,
25:16 --> 25:18and one here in the Northeast.
25:18 --> 25:20And that's kind of mostly it.
25:21 --> 25:26And all of them are at the very least,
25:26 --> 25:30of course Doctors of philosophy.
25:30 --> 25:33So PHD's, but MD's and MD,
25:33 --> 25:35PhDs and MD's who do decision
25:35 --> 25:38science are far and few in between.
25:38 --> 25:39In the United States specifically,
25:39 --> 25:41this is different in Europe
25:41 --> 25:43and different in Canada.
25:43 --> 25:45And that ties into point #2,
25:45 --> 25:47which is that
25:47 --> 25:49in general, you know,
25:49 --> 25:50the decision science umbrella
25:50 --> 25:53includes so many different aspects
25:53 --> 25:55where you can do simulations,
25:55 --> 25:56where you can weigh decisions.
25:56 --> 25:58But if you want to completely
25:58 --> 25:59separate from that,
25:59 --> 26:02you can also layer in costs.
26:02 --> 26:06And I think that is especially here
26:06 --> 26:09in the United States when you start to
26:09 --> 26:11talk about those two concepts together,
26:11 --> 26:14costs and effectiveness, right?
26:14 --> 26:16So cost effectiveness,
26:16 --> 26:17especially during
26:17 --> 26:20the period here in the
26:20 --> 26:23Mid 2000s and the early twenty 10s
26:23 --> 26:25with the Affordable Care Act and
26:25 --> 26:27this conversation about who makes
26:27 --> 26:29decisions about your health care,
26:29 --> 26:32who makes decisions about how much
26:32 --> 26:34is too expensive to pay right.
26:34 --> 26:38These are discussions that in some
26:38 --> 26:42ways shaped and morphed the discussion
26:44 --> 26:47unwillingly in a way about
26:47 --> 26:48about decision analytics,
26:48 --> 26:52but we're in a period where now
26:52 --> 26:56our President has signed into law
26:56 --> 27:01an act that will go forward
27:01 --> 27:04in 2026 and give CMS an opportunity
27:04 --> 27:06to start negotiating drug prices.
27:06 --> 27:09So I think reason #2 has to do
27:09 --> 27:11with this thorny issue of costs
27:11 --> 27:13and who makes those decisions.
27:13 --> 27:14The reality is, at the end of the day,
27:14 --> 27:16cost also matters, right?
27:16 --> 27:19And we need to be able to account for it.
27:19 --> 27:21Now, whether we make decisions on it or not,
27:21 --> 27:23it's totally up to us.
27:25 --> 27:28I mean, one would think that
27:28 --> 27:30decision analytics plays such a key
27:30 --> 27:33role in terms of actually grounding
27:33 --> 27:37the cost decision in data and on risks
27:37 --> 27:41at each decision point along the way.
27:41 --> 27:43You mentioned that you're interested
27:43 --> 27:45in public policy and using decision
27:45 --> 27:48analytics to guide public policy and
27:48 --> 27:51at the same time individualized care.
27:51 --> 27:53Can you talk a little bit in our
27:53 --> 27:55last minute about how those two
27:55 --> 27:58are either at odds or how they come together?
27:59 --> 28:01Well, I think they can fuse beautifully
28:01 --> 28:03together, but methodologically
28:03 --> 28:05they need to stay separate.
28:05 --> 28:08There are definitely ways that we can help
28:08 --> 28:10individuals personalize their treatments.
28:10 --> 28:11And one of the avenues that we're
28:11 --> 28:13going to expand out into is looking
28:13 --> 28:15at out of pocket costs in this realm,
28:15 --> 28:18which hasn't really been done a lot at all.
28:18 --> 28:20And then separate from that,
28:20 --> 28:23keep the health system policy issues separate
28:23 --> 28:25and the stakeholders are very different.
28:25 --> 28:27So you need to be able to cater
28:27 --> 28:28to those specific stakeholders
28:28 --> 28:30and I think we're
28:30 --> 28:31going to be able to do both.
28:31 --> 28:34Doctor George Goshua is an
28:34 --> 28:36assistant professor of medicine in
28:36 --> 28:39hematology at the Yale School of Medicine.
28:39 --> 28:41If you have questions,
28:41 --> 28:43the address is canceranswers@yale.edu,
28:43 --> 28:45and past editions of the program
28:45 --> 28:48are available in audio and written
28:48 --> 28:49form at yalecancercenter.org.
28:49 --> 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.

Information

Classical Hematology with guest Dr. George Goshua November 6, 2022 Yale Cancer Center visit: http://www.yalecancercenter.org email: canceranswers@yale.edu call: 203-785-4095