The Second Century in the Land of Small Tumors

April 26, 2023

Yale Cancer Center Grand Rounds | April 25, 2023

Presentation by: Dr. James Yao

Information

ID: 9872
To Cite: DCA Citation Guide

Download Transcript

00:00Good afternoon, everybody.
00:01I think we'll go ahead and get started.
00:03So greetings to everybody in
00:05the room and to everyone online.
00:08And welcome to Yale Cancer Center,
00:10Grand Rounds. My name is Pam Koons.
00:12I'm the director of the
00:13Center for GI Cancers,
00:14for those of you who do not know me.
00:16And it is a great honor to introduce
00:19my friend Doctor James Yao as the
00:22Norbert Schnog Endowed Lecturer.
00:23So Doctor Yao is a Professor and
00:25Chair in the Department of GI
00:28Medical Oncology at the University
00:30of MD Anderson Cancer Center,
00:32and he received his medical degree
00:34from Baylor College of Medicine and
00:37completed his fellowship at MD Anderson.
00:40So for the last two decades, Dr.
00:42Yao and colleagues have really transformed
00:44the field of neuroendocrine tumors.
00:46So that is how I know him and he has led
00:50practice changing randomized clinical trials,
00:53specifically the family of radiant clinical
00:56trials that include the drug everolimus
00:58that led to FD approvals for pancreatic net,
01:02lung net and GI and undercon tumors.
01:05Doctor Yeah,
01:05I was also a strong advocate
01:07of mentoring and education.
01:08He is a founding member and past
01:10chairman of the North American
01:12Neuroendocrinic Tumor Society of which
01:14I am the president of this year.
01:16And then through that society can helped
01:19establish two young investigator awards
01:21that fund early career investigators.
01:23He's also the past chair of the
01:25NCIA and under consumer task force.
01:27And during his tenure,
01:29early career and female investigators
01:31led more than 50% of the multicentered
01:33clinical trials developed through
01:35that net task force.
01:36I've known doctor Yes since I was a fellow.
01:40I am one of those early career
01:42investigators who benefited from his
01:43mentorship and scholarship and had
01:45the opportunity to lead one of the
01:47randomized trials through the net task force.
01:49So I'm grateful for you coming today
01:52and joining us to speak on the 2nd
01:54century of the land of small tumors.
01:56So thank you,
02:02Thank you so much for that kind introduction
02:05and very glad to be here with you today.
02:07So today I'm going to talk a little
02:10bit about your endocrine tumors,
02:12where we've been and some of the challenge
02:15remains and for you know what we need to.
02:18Make the next century even better
02:20than the what we've done so far and
02:24it's a plug for nanettes this this
02:27rainbow at this photograph was from
02:29one of the nanettes meetings which
02:31we held at the at the the Grand Tea
02:35Towns National Park and let's see
02:42here's my disclosures.
02:44So the field of neuroendocrine tumor
02:47started with open door for first
02:50described this entity about in 1907 he
02:53described this group disease is cancer
02:56like or part of what tumors are more slow
03:00growing than the typical carcinomas.
03:03The 1st century of net has been a a
03:06century where we've learned a lot about
03:08the Natural History of the disease.
03:10Understand a lot of the endocrine
03:13manifestations of neuroendocrine tumors
03:15and we also learn a lot about epidemiology
03:19of disease in semester biology.
03:22However, the number of therapeutic
03:25introduced over this period is
03:28actually relatively sparse prior to I
03:32would say the more recent approvals.
03:35There was only one drug that was FDA
03:37approved for oncologic control and that
03:40streptosis dosen for pancreatic net.
03:43There were two drugs approved for hormonal
03:46control of the neuroendocrine tumors.
03:51This is certainly not for lack of effort.
03:54This is a classic lecture by Chuck
03:57Mortell where he talks about his
03:59odyssey in the land of small tumors as
04:01you can see on the table on the right.
04:03There's been numerous agents that
04:06were studied but these chemotherapy
04:09agents did not really have that
04:11much activity with the exception of
04:15DTIC and streptozosin in pancreatic
04:19neuroendocrine tumors.
04:20Another thing you'll see is that
04:22that because you know this disease
04:24was thought to be rare and he that's
04:27why he used the the the term land
04:30of small tumors.
04:31The studies were actually very small and I
04:35think that really limited the the progress.
04:38These were all single arm studies
04:40and some of them only containing
04:42less than a handful of patients.
04:48So one of the first things I think we
04:50needed to understand about neuroendocrine
04:52tumors is that the disease is probably
04:55actually more common than we think.
04:57One of the analysis we did in from
05:00the SEAR database and we showed that
05:04comparing to other malignant neoplasms
05:07diagnose incidence of neuroendocrine
05:09tumor is continually rising and since
05:12this we have a kind of updated the data
05:15and in you know when when it was in 2004
05:20the incident was about 5 per 100,000,
05:242012 about the.
05:267 to per 100,000 and more more
05:29recent data would happen publish
05:32it is well above 8 per 100,000.
05:35Another thing that's different about
05:37this disease is because the disease
05:39is more slower growing patient live
05:41a lot a lot longer with the cancer.
05:44So essentially the prevalence
05:46statistic which is the number of
05:49patients who are potentially in
05:51need of care because there are life
05:53with disease is actually higher.
05:55So if you did limited duration
05:58prevalence analysis which we did from
06:00the SEAR data it the US prevalence
06:03last we looked was above 170,000.
06:06So certainly this is still at least
06:10for the moment below the 200,000 cut
06:13off which the FDA uses this definition
06:17of rare disease and certainly if you
06:19further divide you're in the consumer
06:22into a subtypes that will remain.
06:24Rare for quite a quite a long time.
06:27So one of the question to think about is,
06:30was this rising incidence and so
06:32forth is what's going on here.
06:35There are environmental factors that
06:37are increasing the incidence of
06:39neuroendocrine tumor or perhaps this is
06:42just better recognition of the disease.
06:45Certainly we are seeing more
06:47neuroendocrine tumors in some case
06:50related in the gastric urine the consumer.
06:53Related to use of PPI's,
06:56but I think for the most part these
06:59neuroendocrine tumor has always been there.
07:01So here are a couple of classic
07:03studies in two in,
07:04in,
07:05in Carcino tumors are really talking
07:08about intestinal neuroendocrine
07:10tumors are two studies that included
07:1215,000 autopsies and these tumors
07:15are found in about 1% of autopsies.
07:18So these are patient who died
07:21from unrelated causes.
07:23And most mostly lift out their natural
07:26lifespan without having them diagnosed.
07:29So really the question is not so
07:31much whether they're increasing in
07:35and what are environmental factors,
07:37but what transforms some of these nine
07:40small tumor into malignant ones but
07:43pancreatic neurin different tumors.
07:44There's one study that was in Hong Kong,
07:47again 11,000 autopsy one in 1000.
07:51Autopsy specimen had a pancreatic urine.
07:53The consumer,
07:54if you look for them compare this
07:57to a diagnosed instance more like
07:59in the range of maybe three to five
08:02per million per year.
08:04That tells you probably less than
08:071% of pancrea and urine.
08:09The consumers that are present
08:12and in patients eventually become
08:15clinically relevant.
08:16This is posing a challenge for
08:18us as we move forward.
08:20Because the increased use of
08:23imaging nowadays you can hardly
08:25go to the ER with abdominal
08:27pain without leaving the ER with a CT scan.
08:30So we're finding a lot of small tiny
08:33pancreatic neuroendocrine tumors,
08:35some of them in the head of pancreas
08:38where if you try to operate on them may,
08:41may, may be a quite a morbid
08:43and higher risk procedure.
08:44So understanding which of these can be
08:47left alone and patients are going to.
08:50Essentially lived with disease
08:52in their natural lifespan,
08:53which one is near to near that
08:55really needs to intervene on is
08:58going to be important going forward.
09:00So the other thing that the with you know
09:04this information about the incidence
09:06and prevalence in your endocrine tumors
09:08is that the patient advocacy groups,
09:12you know in the past decades
09:14has really got engaged.
09:16There are the stories of
09:18patients who have had.
09:19Long history of symptom maybe
09:22that went undiagnosed for decades.
09:24So there's a there was a strive to see
09:27whether we can recognize the symptoms
09:30earlier and diagnose the cancer earlier.
09:33But the challenge is the symptoms that are
09:36associated with these tumors are fairly
09:39vague and common in the general population.
09:42So this is study we did
09:45from CR Medicare database.
09:46Essentially, looked at the year prior
09:49to their neuron cancer diagnosis,
09:51what kind of doctor do they go visit and
09:54what sort of symptoms do they complain of?
09:58You can see, well,
09:59statistically significant for most of these.
10:02There are differences in
10:04rates of hypertension,
10:05abdominal pain, heart failure,
10:08diarrhea, and peripheral edema.
10:09But if you try to look at a positive
10:12predictive value of these symptoms
10:13when you're in the current tumor.
10:15They're all very,
10:16very low because they're very
10:18common in the general population.
10:23The newer endocrine field also is
10:25a field where the very terminology
10:27we or she used to describe the
10:30disease has been evolving in the
10:33in in the over the past decades.
10:35In the older time frame,
10:38the worst like carcinoy eyelid
10:40spells were commonly used.
10:42And it's moved to newer endocrine
10:46neoplasms and there's grading
10:48initially just grade 1-2 and now
10:51differentiation is added to add a
10:54historical context on why the the
10:56constant change almost feels like
10:59in terminology is that this field
11:01you know at the time when these
11:04terminologies classification created was.
11:06Relatively.
11:07I think people didn't really know where
11:09the right cutoff is in terms of the disease.
11:12It's more based on consensus and
11:15recurrence and relate the true biology.
11:18What beginning to understand is
11:21clearly there's two different group
11:23of diseases well differentiated,
11:25you're in the consumer grade 1-2 and
11:28three and they're mostly grade 1-2 and
11:30numerically and then the essentially
11:33the poorly differentiated urine carcinomas.
11:36Which is a completely different disease
11:38that has nothing to do with the other,
11:41right.
11:42And there are also differences
11:44in terms of the primary site.
11:47We'll talk a little bit about the
11:49molecular landscape and genomics
11:51of the different primary sites,
11:53but they are characterized by
11:56relatively low tumor mutational burden,
11:59but these tumor actually has high
12:02high rates of chromosomal instability.
12:05You see instead of point mutations,
12:08a large scale chromosomal changes
12:11lung neuro endocrine tumor.
12:13The most common mutation see is M
12:17EM1 and the same with pancreas M EM1.
12:19But here you also see DAX and ATRX and
12:23intestinal relatively few somatic mutations,
12:28but you see frequent loss of chromosome 18,
12:31the poorly differential
12:32neuro endocrine tumor.
12:34It's probably really a mixed bag a
12:36lot of time these are essentially
12:39transformed versions of adenocarcinoma,
12:42occasionally transformed lower grade
12:44tumor after certain types of therapy,
12:48but they're characterized by a very
12:50fast growth rate and mutation in TP53
12:53and RV are the most common mutations.
12:58So if you understand the genomics
13:00of neuroendocrine tumors,
13:01so one of the things we did is leverage
13:04our large phase three clinical trials.
13:07We did a series of trials called radium
13:10trials looking at everolimus where
13:12over about 1000 patients across you
13:15know four studies were were enrolled
13:17and where we can get the tumor.
13:19We did a whole genome analysis.
13:25We saw relatively few somatic mutations,
13:28but what what is striking is the amount
13:32of largescale chromosomal changes
13:34that you see chromosomal gain and
13:37chromosomal loss and these actually
13:39have very significant prognostic value.
13:42So for example in pancreatic
13:45neuron different tumors,
13:47patient with high chromosomal instability
13:49actually have a much better prognosis
13:52in the advanced disease setting.
13:54And we'll talk about a little bit
13:56in the next few slides why that
13:59is because it is a specific you
14:01know carcinogenesis pathway that's
14:04this this is implying here.
14:06And then we see also those patients
14:10with intestinal neuroendocrine tumor
14:12with loss of chromosome 18 also have a
14:15far better prognosis than those who do
14:18not have a loss of chromosome 18
14:21whereas the loss of chromosome 3.
14:23On the lung neuron, different tumors
14:27pertains to a poor prognosis.
14:30So one of the things that really always
14:33short struck me is really what's going on
14:35with pancreatic neural in the consumer.
14:37It's really one of my favorite
14:39diseases in the sense there's so much,
14:41so much stuff here.
14:43So you see here when we sequence
14:46the pancreatic neural in the tumors.
14:49They roughly fall into three categories
14:51when you look at the host whole genome
14:54in terms of chromosomal changes.
14:57In the first group here, Group One,
14:59they lose one copy of 11 of
15:03the 22 chromosomes.
15:05In the second group,
15:07there's loss of 1 copy of the 11 chromosomes,
15:1211 one copy of 11 chromosomes.
15:14And gain on the complementary 11
15:18chromosomes and then there's a
15:21group that are relatively stable in
15:24terms of chromosomal abnormalities.
15:26And on the bottom panel is little small.
15:28So I'll just talk through it a
15:30little bit and it's important in
15:32the sense that you can actually
15:35link these chromosomal changes to
15:37specific mutations that are present
15:39in if you look at this.
15:42The chromosomal instability tumors
15:44so that that's these are in Gray and
15:48in red are essentially are in rich
15:51for patients with M EM1 mutations.
15:54So what's the link between M EM1
15:58mutation and and this and the M EM1
16:02mutations is also linked with DAX
16:05mutations whereas the ATRX mutations.
16:10Essentially also involved in a TRX and
16:12DAX are involved in alternative links.
16:15Near telomeres can be associated with
16:19chromosome instability in absence of M EM1.
16:23So the ATRX by itself the mutation
16:27seems to drive this phenomenon.
16:30So so So what we see here is then
16:32you see DAX and ATRX mutations
16:36associated chromosomal instability.
16:38And you have,
16:39you know loss of 1 copy of 11 chromosome
16:42and gain on the complementary 11
16:45chromosomes and the strong association
16:47between men one mutation and DAX
16:49in the combination of men one DAX
16:52mutation with chromosome instability.
16:54So what's going on here?
16:56Why are we losing one copy of
16:5911 chromosome and gaining on the
17:02complementary 11 chromosomes?
17:05For whatever reason,
17:06you're essentially what's actually
17:08going on is you're losing one copy of
17:1211 chromosomes and this in some patients,
17:14probably due to happily insufficiency,
17:17is leading to whole genome duplication.
17:20So essentially these are copy neutral LOH.
17:25They are occurring essentially in the
17:29game because the whole genome duplication.
17:32Is occurring in the
17:34complementary 11 chromosomes.
17:36So what's the story here?
17:38While the most common mutation in your
17:42endocrine tumor is man one specifically
17:44linked to pancreatic neuroendocrine
17:47tumors occurring roughly about 40%
17:50of patients and also associated with
17:53lung neuroendocrine tumors.
17:55What do we know about man one biology?
17:59It is certainly is epigenetic
18:01regulators involved in modulating
18:04P27 and it's actually involved
18:07in controlling endocrine mass.
18:09So this is a study done at Stanford
18:13where the group looked at men
18:15and in in mice doing pregnancy
18:19and you can see that men and
18:22expression goes down during pregnancy
18:24and goes back up post pregnancy.
18:27Associated with that is turning on cell
18:32cycle and increase in endocrine mass.
18:35And so there is a, you know there's
18:40important biology here in prevention of
18:43gestational diabetes related to men in
18:48men in turns out is also an
18:52important regulator of telomeres.
18:54In the Nurses in the Prostate, Lung,
18:59Colorectal Ovarian Cancer Screening
19:01Trial and Nurses Health Study that
19:04involved about 3600 patients,
19:06the group this group evaluated 743
19:09snips and try to correlate that with
19:13essentially peripheral blood telomere lens.
19:16The only gene that fell out to be
19:19important was actually men and.
19:21It was the most important implicated
19:24in control of telomere lens
19:27for for in in the study.
19:31So the story of telomeres,
19:34you know as you know the telomeres are
19:37in the caps and end of our chromosomes
19:39and Menon is driving cell cycle in here.
19:43The telomere lens is going to get
19:46short as telomere lens gets short.
19:49Essentially usually the cancer
19:51cell dies where you need to turn
19:54on some way of maintenance of
19:56telomere or Linston telomeres.
19:58For most cancers this is essentially
20:02activation of telomeres,
20:03but in a few cancers and in in
20:07pancreatic neuroendocrine tumors,
20:08the mechanism that's gets gets gets activated
20:13as alternative linsing of telomeres.
20:17How do we know this?
20:18This is some slice courtesy of
20:21Christopher Heefy where he showed
20:24essentially in neuro endocrine tumor
20:27that has well typed Dax ATRX you see
20:31fairly normal telomere lens and when
20:35there is Dax or a TRX alterations
20:39you see these bright pink spots
20:42which are telomere specific fish.
20:45Showing a classic pattern associated
20:50with alternative listening of telomeres,
20:55the story on essentially alternative
20:59listening telomeres and DAX ATRX mutations
21:02is actually complex in terms of prognosis.
21:06Earlier on I showed you a slide
21:09where essentially the the.
21:11Mutation of DAX ATRX and and turning
21:17out ELT was associated with good
21:19prognosis in patients with advanced
21:22pancreatic neuroendocrine tumor.
21:25The situation is actually
21:27reversed in the earlier disease.
21:29Essentially what's going on is that
21:32advanced disease the the DAX ATRX mutation.
21:38Is marking a group of pancrea urine
21:41the consumer who goes down a very
21:44specific carcinogenic pathway,
21:46whereas in in the earlier disease
21:49this actually the IT pretends
21:51to be a worst prognosis.
21:54So this is a great study
21:56that was done in men,
21:57one families.
21:58So these are patients with
22:01familial mutations in M EM1.
22:04What they're able to show is that.
22:07When the tumors are small,
22:09you usually don't see DAX ETRX
22:13mutations and the DAX ETRX mutations
22:16occurs in tumors that are larger
22:20in this case and I think they
22:22use a cutoff about 3 centimeters
22:24and also happens in patients
22:27who have lymph node metastasis.
22:29So likely what's going on is
22:33that as the these tumor.
22:35Are driven by men one to
22:38proliferate these benign tumors.
22:41The tilar mirrors are getting shorter
22:44and the ones who are able to turn on
22:47tilar mirror maintenance throughout
22:49are the ones that gets larger and
22:52then lead to regional metastasis.
22:59So again, this is just showing
23:02the same in terms of A.
23:05In a recurrencefree survival graph,
23:08those who are turning on health
23:10in the localized setting where
23:12they have three section have
23:15a little bit poor prognosis.
23:21Next I'm going to shift gear a little
23:22bit and talk about essentially on the
23:25clinical side the development of new
23:28novel therapies for neuroendocrine tumors.
23:31So essentially prior to 2007,
23:35we only had Streptozosin for your
23:37in the contumor of the pancreas.
23:40And since then you really have
23:42seen a lot of new agents showing
23:45activity getting FDA approved for
23:48having positive phase three trials.
23:51And I think a key thing here that
23:55happened really related to one of the
23:58meetings in Pam you were involved with.
23:59Was the first in a CTPM meeting sponsored
24:03by NCI and the importance of that
24:06meeting is really to come to consensus.
24:09What is the right kind of clinical trial
24:12design when you're in the consumers,
24:14what are the correct endpoints?
24:17There's a recognition progression,
24:19free survival is probably the
24:21right endpoint or in,
24:23but the phase three trials
24:26are are recommended.
24:27Overall survival trials,
24:29neuroendocrine tumors,
24:31we came that out in doing the meeting
24:34and realized they will require a
24:36probably about two to 3000 patients
24:39and probably 8 to 10 years to execute.
24:43So that that's why you will see
24:47in the subsequent slide most of
24:50the approved agents are able to
24:52then demonstrate PFS benefit.
24:54But we don't have quite a
24:56large sample size needed that
24:57demonstrates survival benefit
25:02going going into the the systemic
25:04randomized space free trials are you
25:07know we you're going to talk to them
25:09a little bit about different targets.
25:12So the first targets we'll talk
25:14about is the Smestan receptor.
25:16For a long time prior to this Smestan
25:19receptor targeting was Octreotype.
25:22Was approved for control of Carson
25:24syndrome it relief flushing and diarrhea
25:26in probably about 70% of the patients.
25:29But there are a lot of back and
25:31forth debate as to whether actually
25:33or not is slow cancer grows and it
25:36was almost like a little religion
25:39people either believe it or we they
25:41they didn't but what was important
25:43is you just need to actually do the
25:45trial it turns out and in this phase
25:48three trial that's done by the.
25:51A multicenter German trial in in patients
25:56were relatively newly diagnosed with
26:00small bowel neuroendocrine tumor.
26:02They were able to demonstrate
26:04improvement progression free survival.
26:06A similar trial was land Realty
26:09was conducted as a larger trial
26:12and included a broader group of
26:14patients including pancreatic and
26:16rectal neuroendocrine tumors.
26:18And again showing significant benefit
26:20in terms of progression free survival.
26:23Notice however the hazard ratio
26:25for the octerotized study was a
26:28little bit lower than the hazard
26:30ratio for the land real time study.
26:33This is probably a by byproduct in terms
26:37of the way the trial were executed.
26:41It turns out the octerotized study
26:44was permanent terminated early.
26:46At interim analysis and in there's
26:50been subsequent publications and
26:53analysing analysis of popular population
26:55of studies that can demonstrate
26:58while when you terminate a study
27:01early for outstanding efficacy,
27:03you tend to overestimate the the
27:06magnitude of the treatment effect.
27:08And that's just a you know a byproduct
27:12of our early termination because
27:14when you terminate a study early.
27:16You preserve your ability to
27:19test the hypothesis,
27:21but not the ability to estimate
27:23the magnitude of treatment benefit.
27:27Another way to term to target
27:30some mass and receptor is PRRT,
27:33which really has become very well,
27:36you know, widely used at this point.
27:39Again, in the earlier
27:41development of PRRT it was not.
27:43It was a lot of a single single
27:48institution studies and you have
27:50these publications in high impact
27:52journals where they purportedly report
27:55a phase two study of 1000 patients.
27:59And you know,
28:00but actually what was needed for
28:02really demonstrating benefit and
28:03approval is a randomized phase three
28:06trial which you can do actually was
28:08far fewer than thousand patients.
28:11So this takes advantage of the
28:13fact that semastin receptors are
28:15present on your endocrine cancer
28:17cells in 7080% of the cases.
28:20Specifically for semastin receptor
28:222 when the lichen binds to
28:25the receptor is internalized.
28:27So.
28:28So these agents essentially takes a
28:32Lutetian 177 and taking into the cell
28:35and leading to very good efficacy.
28:41There's also a role for targeted
28:44therapy in neuroendocrine tumors.
28:46One of the drugs that we were involved
28:49in developing is everolimus of
28:51affinitor targeting the emptor pathway.
28:54The Radian 3 trial was the first
28:57to report out and for pancreatic
29:00neuroendocrine tumors and here you saw
29:03a benefiting progression free survival
29:05from median 4.6 months to 11 months.
29:10And hazard ratio was .35 here
29:13in overall survival because the
29:15crossover we did our PFST analysis
29:18rank preserving structure failure
29:20time showing like there's a likely
29:22benefit in overall survival,
29:24but in because of the the crossover
29:27these such studies and not these
29:30studies are really designed
29:31to evaluate overall survival.
29:37For Radian 4, this is the phase
29:39three study we did in lung
29:42and GI neuroendocrine tumors.
29:44Again patient were randomized to
29:47receive everolimus or placebo.
29:50The the the, the PFS improved from
29:543.9 months to 11 months with a
29:57hazard ratio of 0.48 and a trend
29:59to our overall survival benefit.
30:05Another targeted agent that's
30:07shown benefit is sunitnip.
30:09So Sunita was initially evaluated in a
30:12phase two study being that had two cohorts
30:15for intestinal neuroendocrine tumors and
30:18for pancreatic neuroendocrine tumors.
30:20All the responses were seen in the
30:23pancreatic neuroendocrine group.
30:25So it was taken into a phase three trial.
30:28The study actually terminated early
30:31at an unplanned interim analysis.
30:33Nonetheless there it was significant
30:36benefit demonstrating PFS and then that
30:39led to the FDA approval of the drug
30:43for pancreatic neuroendocrine tumors.
30:47We do believe VEGF inhibitors may
30:49have a role in extra pancreatic
30:52neuroendocrine tumor as well.
30:54This is a another phase three trial
30:56that I did early in my career,
30:59the SWAG O 518.
31:01And the in this study patients were
31:05randomized from octreotype plus interferon
31:08versus octreotype plus Bebasus MAP.
31:11Where we're able to show in this
31:13study is that although the response
31:15rate improved with Bebasusan MAP and
31:18toxicity was better was Bebasusan map,
31:21there was not any significant
31:24difference in progression free survival.
31:27So this is probably one of
31:28my regrets in the career.
31:30I probably should have done this
31:32study against placebo and we would
31:34have had had another drug available
31:36for neuroendocrine much earlier on.
31:40This is what the time point in my
31:43career where we weren't sure whether
31:45we can execute a placebo control trial.
31:47It's certainly a little bit harder to do,
31:49but often placebo control trial
31:52give you cleaner data.
31:54Especially when the comparator
31:56arm is not is not very carefully
32:00what is not well defined.
32:04So there has been others who evaluated
32:08veget inhibitors in your in the Contuber.
32:12This is a study conducted
32:14also in the cooperative group.
32:16The Pi is Emily Burksland and
32:19patient were randomly assigned to
32:22either pizopanit versus placebo.
32:24And there there was the benefit
32:26in terms of progression free
32:29survival also demonstrated in extra
32:32pancreatic neuroendocrine tumors.
32:34So potentially showing the importance
32:37of role of VEGF inhibitor outside
32:41beyond the pancreatic group in terms
32:45of phase three studies for extra
32:49pancreatic neuroendocrine tumor.
32:51And there's also a study that was
32:54performed in two studies that were
32:57performing in China with Serofatin NIP,
33:00another VEGF or multi kinase
33:03inhibitor demonstrating similar
33:05magnitude of benefit for Serofatin
33:08Nip both in pancreatic net and extra
33:11pancreatic net unfortunately the FDA.
33:13It's going to probably require
33:16the the company to redo the trial
33:19because it did not contain it was a
33:22purely Chinese population and the
33:24population may not fully represent
33:26the lines of prior therapy Western
33:29populations would have been exposed to.
33:34Next I'm going to mention while
33:36Doctor Kunz's trial Ecog 2211,
33:39this is actually a very important trial.
33:43Partially because the initial development
33:46of Timosolomite were essentially
33:48skipped the single agent step they were,
33:52you know most of the trials that
33:55were published were doublets.
33:57So always been a question to feel
33:59that whether you need doublets
34:01or you know where the agent is,
34:05Timosolomite by itself is a sufficient.
34:08Certainly there's rationale to
34:09look at this class of agents in
34:12pancreatic neuro in the consumers.
34:13If you dig back into Chuck
34:16Mattel's papers and so forth,
34:19DTIC is active in the disease.
34:21So this is a trial that compared Timosolomite
34:25to Tim Cape at the intern analysis.
34:28The study met its primary
34:31endpoint and showed improvement
34:33in progression free survival.
34:35For our patients with Tim Kay and I
34:39think another actually very important
34:42finding from this study is the
34:45prognostic and significance with
34:47association of the MGMT expression
34:51with the response in this is a
34:55DNA repair pathway when that often
34:59are methylated MGMT and leading
35:01to low expression and you can see.
35:03That for patients with low MGMT,
35:06the response rate is much higher than
35:09those who have intact MGMT expression.
35:14So if you look at the current treatment
35:16landscape for neuroendocrine tumor,
35:17we have come a long way.
35:20You know in the beginning historically we
35:22only have one agent for pancreatic net.
35:25Now you have number of phase three
35:28clinical trial covering many of the
35:31different in your endocrine tumors.
35:34Essentially these are clustered
35:36around agents that targets these.
35:39These are stable or early disease like
35:42TRILTY and then Realty in the pro
35:45Med and the CLARINET study and in the
35:49studies who tend to target patients
35:52was faster progressing disease PRT
35:55somewhere in the middle that required
35:58progression in the past three years.
36:00And most of the targeted agents
36:02require progression in the past one
36:05year when in the case of Radian 4
36:07progression within the past six months.
36:10So what are some of the remaining
36:13challenges and questions that we
36:14we have when you're in the current
36:17tumor at this point,
36:18one of the question I get asked
36:20the most is sequencing,
36:22what's the optimal sequencing of
36:25therapy for neuro in the current tumors?
36:29So it's kind of interesting
36:31because you're in the consumers,
36:33you had approval a lot of agents
36:35while in a short period span of time.
36:38So they were not really developing a way
36:42where they were specific align first line,
36:44second line, third line.
36:46Most the drugs were either approved
36:49for progressive disease or they
36:51were just approved for advanced
36:54disease but optimal sequencing.
36:56It's really talking about which sequence
36:59leads to the best overall longterm survival.
37:03This is actually extremely
37:05difficult question to answer.
37:08It's not about which agent when used
37:11first has the longest initial PFS,
37:14because if that agent,
37:16you know,
37:17essentially takes out your kidney or makes
37:21it difficult for you receive other agents.
37:24And it may not be the best agent
37:26to use initially.
37:28So almost certainly this is if you
37:30really want to answer this question,
37:32it needs overall survival endpoint.
37:35Well,
37:36here's the challenge right when for
37:39different indications you have different
37:42number of treatments available,
37:45the approved therapy for lung,
37:47there's only ever limus in peanut you
37:49have six agents that are available,
37:52approved you can use.
37:54A 7th agents demonstrated activity
37:56that that's probably works well.
38:00You can imagine trying to compare
38:04optimal sequences.
38:05There's 5040 sequences,
38:095040 arms for overall survival.
38:12This is not where we want to spend
38:14our energy and because I think
38:17likely before evening to solve a
38:20simpler question before you actually.
38:23To answer the question and complete a trial,
38:25the treatment landscape would
38:26have changed in the trial design
38:28will probably no longer be valid.
38:33And to give a actually example
38:34of attempt to do this,
38:36our European colleague
38:38contacted the secretor trial.
38:41The secretor trial look look to
38:44compare the sequence of Ever Linus
38:46followed by Streptozosin based
38:48chemotherapy or Streptozosin based
38:51chemotherapy followed by ever Linus.
38:53They weren't going to be quite ambitious
38:56to try OS as the primary endpoint.
38:58They were going to look at P FS2.
39:01So initially are due to a cruel
39:05issues that they had to do scale
39:08back their ambitions to look at
39:10P FS1 as the primary endpoint.
39:13So what did the study show?
39:16Yeah, actually showed that although
39:19Streptozosin set of toxic chemotherapy.
39:22Was a little bit more toxic but higher,
39:24had a higher response rate,
39:26but there was no difference in progression
39:29free survival between the two arms.
39:31So higher higher response rate may not
39:34necessarily lead to a better outcome.
39:38The second most frequent question I get
39:41asked about nearing the consumer these
39:43days is precision medicines and biomarkers.
39:46If you did a search on your end,
39:49the consumer and biomarkers on Pub Med.
39:52And you'll get thousands,
39:54probably near 10,000 results back.
39:57So what do we know about biomarkers
40:00for neuroendocrine tumors?
40:03I usually think about biomarkers
40:05as two classes.
40:06These are prognostic identifying
40:08those people who have a better
40:11or worse outcome and predictive
40:13meaning to actually sorting out
40:16individual who are more likely.
40:19But then similar individual without a
40:21biomarker to experience a favorable
40:23unfavorable benefit from an exposure to
40:26a medical product we environment agents.
40:28So the bottom line is who should get
40:31this treatment is really the important
40:34question for predictive biomarker.
40:36Another way to think about the importance
40:39of predictive biomarker is really
40:42thinking about like who's going to
40:44benefit from treatment if you have a
40:47treatment where everybody benefits.
40:50Predictive biomarker can almost becomes
40:52essentially a prognostic biomarker.
40:54It's probably of less clinical
40:58importance in the situation where
41:00half the patient will benefit.
41:02A predictive biomarker is super useful
41:06and it's even more important when a
41:08smaller group of patient have profound
41:11benefit, but most people don't.
41:14So what is actually the situation
41:16you are in different tumor which of
41:18these waterfall plot do we look like?
41:21Fortunately it looks like this whereas
41:23a most the patients benefiting from the
41:28treatment within their treatment indications.
41:34So the challenge of predictive
41:37biomarker is essentially you have to
41:39randomize more patient all patients.
41:41Including patients who doesn't have
41:44the biomarker because without that
41:46randomization is very difficult to
41:49understand which biomarker is important.
41:52You should do this when the marker
41:54is suspected to be predictive but not
41:57proven and you have reliable assay
42:00methodology and cut points and there's
42:03reason to expect benefit potentially
42:06in biomarker negative patients.
42:11Much more common we seeing oncology these
42:14days is this approach which is establishing
42:18a efficacy of biomarker population which
42:21means we only essentially randomize
42:24the biomarker positive population.
42:27So here you can prove the biomarker positive
42:30benefit patients benefit from new treatment.
42:34But it's best used when no benefit is
42:36expected in bowel marker negative population.
42:39You don't have any information gained
42:41about the bowel marker negative population.
42:47But often sometimes we get it wrong, right.
42:49We we don't initially fully understand this.
42:52The classic example in
42:54colon cancer is cetuximab.
42:57The initial FDA approval in
42:59clinical trial was for patients.
43:01Who had e.g. Fr expression on I HC?
43:06Turns out that has nothing to do
43:08with whether someone benefits from
43:10situximab or not in colorectal cancer.
43:14And the net example is really kind
43:16something I kind of lived through.
43:19After we started a phase three trial,
43:21a publication came out in science
43:23showing about 15% of the patients with
43:27pancreatic net at M Tor pathway mutations.
43:30So I I I will,
43:32I would gladly admit I was a very lucky
43:35not to know that when I started the trial.
43:38But because it turns out you
43:42know extra pancreatic net,
43:44none of the patients have
43:46mtor pathway mutations,
43:47quote mtor pathway mutations,
43:49but they all benefited from the therapy
43:52and even in the pancreatic net group
43:55those who had mtor pathway mutations.
43:58And didn't have M Tor pathway mutation
44:01have similar magnitude of benefit.
44:03That's not to say that it's not
44:06correct that you know what what
44:09was published is just means that
44:11I don't think we may sometimes
44:13know the full M Tor pathway or
44:16how these drugs actually work.
44:21Those are biomarkers.
44:23In neuroendocrine trials,
44:25the question often is asked about the
44:28semester and syntacriphy in for semester.
44:32And like octreotile and Realty,
44:35the prominence study actually
44:37allowed for both semester and
44:41receptor syntacrification of.
44:45And clarinet study only treated patient
44:54for semester and receptor positive.
44:58This one comes close to a predictive
45:01biomarker which is the degree of
45:03uptake and response and tumor
45:06shrinkage in in in for treatment
45:10with a peptide receptor radiotherapy
45:13as you can see that comparing
45:15to the using the craning scale.
45:17As the expression goes up,
45:19the response rate increase compared
45:22for peptide receptor radiotherapy.
45:26Another biomarker that was evaluated is
45:30is more like a pharmacodynamic biomarker.
45:34In early studies and single arm study
45:36it looked like those patient who had an
45:39early drop in pomegranate had a benefit
45:42for patients treated was everolimus.
45:45But this turned out actually not to be
45:48that useful when we took it to phase
45:51three because the placebo patient
45:52who had a 30% dropping from Granny
45:55A also had a better outcome as well.
45:58So likely this is pointing out some
46:01issues with the assay performance and
46:04also whether we're not you actually
46:05have to test these patient multiple
46:08times before you get a reliable results.
46:12Another BOW marker attempting to look
46:14at the predictive bow marker in terms of
46:18response is looking at profusion CT in
46:21patients treated with veg inhibitors.
46:23We're able to show that in patients treated
46:26with Bebasusan app is open in a flipper
46:29set that essentially baseline parameter
46:31and change after treatment correlated
46:34with the degree of tumor shrinkage.
46:37I think what we learned here is that
46:40these are very difficult to do.
46:42And very operator dependent.
46:44So it's was possible to do it in
46:47clinical trial taking it out to the
46:50wider clinical practice is challenging.
46:52So if you look at a biomarker
46:55landscape for neuroendocrine tumor,
46:57you see that in terms of understanding
47:01whether the treatment work in the in,
47:05in the indication we do pretty well.
47:08Whereas predictive biomarkers,
47:10there are a few promising ones
47:13like printing scale for
47:18PRT&amp;MGMT for Timosolemai,
47:20However, we're still need a lot,
47:23lot more work to do in terms of
47:26getting real predictive biomarkers.
47:30So I mentioned earlier that we
47:32have a lot of approved therapy
47:34but most of these trial were not
47:36designed to ask a survival question.
47:39So you know has all this work been
47:42approval our patients doing better,
47:45we can look back into the SEAR
47:47database again and showing
47:49that the trend in improving,
47:51improving overall survival in
47:53patients with great one to two
47:56metastatic neuroendocrine tumors.
47:58Suggesting that what we did actually
48:01does actually make a real impact.
48:04So next what do you think we need
48:06to do to continue the progress
48:09in your endocrine tumors?
48:11I think clearly we one thing
48:13we learned is the use of robust
48:16randomized clinical trials and we
48:18shouldn't be shy about using placebo
48:21control trial in the right setting.
48:23We do need better availability of neuro
48:25in the model for translational research.
48:28I think we have a baseline group of
48:31therapy that works now to find the
48:33next pathway to target the next target.
48:36I think the the neuron models in
48:39the lab will really benefit us
48:42and we need to obviously explore
48:46novel therapeutic approaches.
48:47I'll just have two more slides
48:50on the modeling part so.
48:53There's a real challenge with
48:55developing models for well
48:56differentiating your endocrine tumors.
48:59There's been many attempt to generate
49:02cell lines, xenographs and organize.
49:04Principally they are limited by a
49:07slow growing nature of the tumor.
49:09So if you think about it in
49:12placebo arm of clinical trial,
49:14you see these tumors takes about
49:18somewhere between 5 to 18 months median.
49:21To show about a 20% increase in diameter,
49:24if you really had a representative Model 1,
49:28those models are very difficult
49:30to keep alive.
49:31Second, will take you years to run
49:33one single experiment in the lab.
49:35So it's it's very,
49:36very, very challenging.
49:37There are models out there,
49:40but many of them are altering in such
49:42a fundamental way that I don't think
49:45they represent your end of biology.
49:47So if you look at the published
49:50cell lines and.
49:51In in the the models out there,
49:55many of them highlighting yellow,
49:57have mutation that do not occur naturally.
50:00While differentiating Nets with P53 and RB.
50:05The remaining usually are
50:07unknown in terms of P53RB status.
50:11So here's the conundrum.
50:13You need a model that's grows
50:16fast enough to actually take.
50:18And can generate enough material
50:20that you can actually do experiments,
50:23but you still need to represent the the
50:27neuroendocrine slow growing biology.
50:30So how do we tackle this?
50:32One of the efforts we've been doing
50:35in our lab is using a genetically
50:37engineered patient derived organo way models.
50:40So what are we doing here?
50:42We know that if you alter P53 or RV.
50:45These these things will grow
50:47and take and proliferate,
50:49but you don't want the P53RB L you
50:52know altered when you're testing,
50:54studying new drugs or understanding the
50:59the biology of Nets.
51:01So we are using a lenty viral
51:04vector to introduce essentially
51:07doxycycline inducible alterations
51:10in key proliferation pathways.
51:13So the idea is you.
51:16Essentially putting a growth on
51:18and off switch into the patient
51:21tumor samples and then you control
51:24it with in this case doxycycline.
51:28We're using either SV40 large T antigen
51:32behind a the behind the promoter or
51:38a altered P53R273 because the P53.
51:42Acts as a tetramer when even one
51:45copy is is actually mutated.
51:47Is in Paris its function
51:51still a lot of work to do to
51:53figure this out because there
51:55are many different variations.
51:56You can do this.
51:58You can directly in fact the primary cells,
52:01you can grow them in organizing.
52:03In fact the organize and what's the right
52:07condition and how to solve this work.
52:10We're happy to show that
52:11we're making progress,
52:12that we can actually use the system to
52:15make your endocrine tumor organoids grow
52:20because the previous attempts to organize,
52:22while you can use growth
52:24factor to keep them alive,
52:25they don't really grow.
52:26So the only way you can study them is
52:29to have a constant stream of material
52:31coming from the operating room.
52:34But each time is a little bit different,
52:36right? So we're hoping that.
52:40Over the next month to year to fully
52:42characterize all the these organized
52:45that we're developing in terms
52:47of what what is staying the same,
52:49what is being altered and to what
52:52extent we can reverse the P53
52:55and and and SV40 induce changes,
53:00we show all the doxycycline to do
53:04drug screening or study the biology.
53:07So I'm going to end the talk here and
53:09maybe just a few minutes for questions.
53:11You
53:25can say it there. Any questions
53:27from the audience or from or online,
53:30maybe one, see if anyone in the chat,
53:34I'll, I'll ask the first question,
53:36what are you most excited about from a
53:39therapeutic standpoint in the next decade?
53:42That's a tough question.
53:45You know I I I think there's still a role
53:48for immunotherapy but probably not with
53:52existing checkpoint inhibitors but maybe
53:56within except for maybe subpopulations.
53:58That's one of the things we're
54:00learning is although tumor tumor
54:03mutational burden is generally low,
54:06you're in the patients live a long time.
54:09So that tumor mutation burden
54:11actually may change over time.
54:13If you look late in the course of disease,
54:16you may find you know patients will
54:19benefit from those those sort of
54:21treatments especially interesting is
54:23like your work with Timozola mine right,
54:26because these are Asians that tend
54:28to induce tumor mutations and and may
54:31increase tumor orientation will burden.
54:34So that's actually I think the
54:36relevant sequencing question.
54:37When you use that early,
54:38does that mean later on they have a
54:40high T MB and you can go back with I/O,
54:43that sort of things?
54:43Yeah.
54:43Kevin,
54:47thanks. Doctor Yellen,
54:49can you comment given the expanding
54:52armamentarian and systemic agents
54:54where you see the evolving role
54:58of surgical therapy fitting in?
55:00It's complicated historically?
55:03Depending on where you are
55:05and who you work with,
55:07what the landscape is and we have
55:11felt like surgical cyto reduction
55:15has a role in this disease.
55:19But I I don't know that that's
55:22as much the case the current
55:24era or it will be in the future.
55:26I think that's the great question.
55:28I think there's still be a very
55:30important role for cyto reduction
55:32in surgery in this disease.
55:34If nothing,
55:35it actually gives the patient a
55:37potentially a long treatment free
55:39interval from systemic therapy and
55:41although the time course here is long,
55:44so metastatic small bowel patients
55:47are living 8 to 10 years.
55:50But if you ask the patient they will
55:52say 8 to 10 years is not enough,
55:54right?
55:54So I think there's still room
55:56to use more modality including
55:58surgery and international
56:00radiology technique and so forth.
56:05The surgery of symptoms,
56:08it certainly can mean, yeah,
56:12you know there's many different
56:14ways it can in you know some
56:17cases patients have essentially
56:19abdominal discomfort from a local
56:21tumor with nodes and the surgical
56:24resection will bypass can be very
56:27important for them even though
56:29palliative and patients who have.
56:31Severe Carson syndrome sometime
56:33refractive therapy and benefit
56:35from the bulking all types
56:40should there's an increase in
56:42incidence but also survival.
56:43Are you able to kind of differentiate
56:46increased diagnosis of otherwise
56:47ability versus advances in therapy
56:50or you know parse this out? Yeah.
56:53So I think one of the ways we're looking
56:56at the survival changes is limiting our
56:58analysis to those with metastatic disease.
57:01There are still very much some limitations
57:03when you look at that sort of data.
57:05But I think the large registry is
57:08probably still the best way to look
57:10at the survival data because when
57:12you look at individual institutions,
57:14you have a lot of referral bias.
57:16You know, those patients who
57:18has surgery are cured,
57:19they don't come to tertiary centers, right.
57:21They're going on and living
57:23their normal lives.
57:24And so the large registry still
57:27have a very important role there.
57:29And the increase in incidence is happening
57:32in distinct areas like rectal is you
57:35know because the screen colonoscopy
57:37you're finding lot of tiny rectal,
57:41you're in the consumers which also
57:44is linked to specific race and
57:47ethnicity issues and in in in small
57:49pancreatic urine the consumer is
57:51going to be something we're going
57:52to have to deal with just the
57:54increase CT start getting done.
57:59Well, thank you so much
58:00Doctor Yao for coming today.