Yale Advancements in Oncology: Implementing the Best Science from ASCO® 2024 and Beyond: Part 1

October 02, 2024

8:00am Welcome and Introductions

Eric Winer, MD, and Roy S. Herbst, MD, PhD

8:15am Unraveling the Genetic Tapestry: Advances in Germline Testing

Veda Giri, MD

8:40am Genitourinary Cancers

8:40am Hot topics in 2024: Joseph Kim, MD

8:55am Case Discussion: Moderator: Dan Petrylak, MD

Presenters: Michael Hurwitz, MD, PhD; Matthew Austin, MD; Yi An, MD

9:20am Gastrointestinal Cancers

9:20am Hot topics in 2024: Jill Lacy, MD

9:35am Case Discussion: Moderator: Pamela Kunz, MD

Presenters: Laura Baum, MD, MPH; Jeremy Kortmansky, MD; Sajid Khan, MD, FACS, FSSO; Kevin Du, MD, PhD, MSCI

10:15am New Tools in the Toolbox: Surgical and Radiotherapy Advances at Yale

Ehud Mendel, MD, MBA; Henry Park, MD, MPH

10:40am Lung Cancer

10:40am Hot topics in 2024: Sarah Goldberg, MD, MPH

10:55am Case Discussion: Moderator: Roy Herbst, MD, PhD

Presenters: Anne Chiang, MD, PhD; Michael Grant, MD; Henry Park, MD, MPH; Andrew Dhanasopon, MD, FACS; Yu Zhang, MD, PhD

11:10am-12pm Rapid Shots - Practice Changing Literature Across Disciplines

11:20am Melanoma – Harriet Kluger, MD

11:30am Neuro-Oncology – Nicholas Blondin, MD

11:40am Gynecological Cancers – Elena Ratner, MD, MBA

11:50am Head & Neck Cancers – Barbara Burtness, MD

12:00pm Palliative Care – Elizabeth Prsic, MD

1:00pm Keynote Lecture and Q&A Discussion

Artificial Intelligence in Oncology: Challenges and Coups

Lucila Ohno-Machado, MD, MBA, PhD

1:45pm Hematologic Malignancies

1:45pm Lymphoma Hot topics in 2024: Tarsheen Sethi, MD, MSc

Information

ID: 12155
To Cite: DCA Citation Guide

Download Transcript

00:01Medical oncology. And we're also
00:02gonna hear about hematology,
00:04one of our our biggest
00:05areas, and we have a
00:06whole section of hematology.
00:08We have many of our
00:09faculty who will be here
00:10during the day.
00:11As you know, we're out
00:12at fifteen plus fifteen plus
00:14centers around Connecticut and Rhode
00:16Island. So we really, hope
00:17that this will be a
00:18great day to assimilate the
00:20data, talk about how we're
00:21using the data. I know
00:22I always find this,
00:24useful in
00:25in needing to take the
00:26boards again in a year
00:27or two. This is always
00:28helpful to me. So this
00:29is, I think, the thirteenth
00:31time I've hosted this, but
00:32this actually was begun even
00:34before me, by Eddie Chu
00:36when he was in charge
00:37of, oncology. But it's just
00:39great that we have this
00:40this meeting every year.
00:42I'm going to make my
00:43remarks brief. I'll be moderating
00:45a little bit later. We
00:45have a a great day,
00:46by the way. We're gonna
00:47hear from not only most
00:48of the major tumor types,
00:50solid and liquid, but we
00:51have a very special lecture
00:52at lunchtime
00:53by Lucila Ono Machado, who's
00:56our department here now of
00:58informatics,
00:59bioinformatics, and and and she's
01:01gonna really talk about how
01:02we're using big data, AI.
01:04I'm sure everyone thinks about
01:05that a little bit. Your
01:06phone probably does it. So
01:07it's really gonna be a
01:08great day. But now it's
01:10my true pleasure to introduce
01:12our director,
01:14I think known to most
01:14of you, doctor Eric Warner.
01:16And he's gonna say a
01:17few, open remarks and then
01:18turn it over to Pam
01:19to start the session. Eric.
01:27Thanks, Roy. I'll start my
01:29opening remarks by saying that
01:31I always feel like I'm
01:32about to kill myself going
01:34from ninety one to ninety
01:35five.
01:36And this morning was no
01:38exception.
01:39And
01:40as I was racing to
01:41get here,
01:42I thought, good god. I
01:43hope these cars avoid me.
01:45And the they're just it's
01:46a little frightening.
01:49But, in any case, I've
01:50been here for two and
01:51a half years. I'm slowly
01:52getting used to the roads,
01:55and, and it's really a
01:56pleasure.
01:58This used to be called
01:59an ASCO review. Technically, it's
02:01not an official ASCO review.
02:03It's a review of our
02:05interpretation
02:06of of
02:07of,
02:08some
02:09ASCO abstracts.
02:10But beyond that, it's talking
02:11about advances in oncology
02:13and talking about how we
02:14take care of patients. And
02:15in my mind,
02:17meetings like this are very
02:18important for a number of
02:20reasons.
02:21Perhaps first and foremost, it
02:22gets a whole group of
02:23people together,
02:24and there's real value
02:26to that.
02:28And I think we've all
02:28learned that particularly
02:30post pandemic.
02:32But apart from
02:34the review of recent data,
02:36we also are gonna spend
02:37a fair amount of time
02:38reviewing cases
02:40and building consensus.
02:42And
02:43certainly across our entire network,
02:46we try very, very hard
02:47to have a consistent approach
02:49to every malignancy
02:51at every site.
02:52And I think meetings like
02:54this really help contribute to
02:55that. So I want to
02:57invite all of you to
02:58actively
02:59participate in the day. It's
03:01really important to have an
03:02engaged audience. Don't be shy.
03:04Speak up,
03:06and,
03:07and take advantage of all
03:09the people who are here
03:10all day. And with that,
03:11I will say thank you
03:13for coming and,
03:14introduce Pam Coons,
03:16who is,
03:17the,
03:19the
03:20the chief of,
03:22yeah, you've got so many
03:24titles. It's like I'm I'm
03:25deciding between the titles,
03:28who's the director of our
03:30GI center of excellence and
03:31the chief of GI oncology,
03:34and,
03:35and just a
03:37remarkable
03:38GI oncologist who leads a
03:40great group. So, Pam, thank
03:41you very much for being
03:42here.
03:48Alright. You redeemed yourself. That's
03:50good.
03:50Good morning, everyone. Welcome. I
03:52am hosting the first few
03:54sessions this morning, so it
03:55is my great pleasure
03:56to introduce,
03:58my friend and colleague, doctor
03:59Veda Geary, to give the
04:00first talk. She will be
04:02speaking on unraveling the genetic
04:04tapestry, advances in germline testing.
04:06Veda has been here for
04:08almost two years
04:09and is the division chief
04:11of clinical cancer
04:13genetics
04:14and the co director of
04:15the early on cancer. Thank
04:18you.
04:39Thank you very much. Good
04:40morning, everyone.
04:42It's it's such, you know,
04:44really great to be here,
04:45and, I'm honored to kick
04:46us off. So,
04:48this is a really exciting
04:49topic about unraveling the genetic
04:51tapestry,
04:52advances in germline testing,
04:55and bringing some of the
04:57highlights from the ASCO twenty
04:58twenty four meeting today.
05:01And so please keep in
05:02mind that this is just
05:03a few of the abstracts
05:04that were presented. There was
05:05a spectrum of abstracts, and
05:06I'll give you a little
05:07bit of the summary,
05:09of this information.
05:10So I have no disclosures
05:12relevant to this presentation.
05:14Just to set the stage
05:16about,
05:17why this topic is brought
05:18forth is I think
05:20we all know that germline
05:22testing, which is hereditary cancer
05:24genetic testing,
05:25is really growing an impact
05:27in oncology,
05:28across the spectrum, whether it's
05:30in cancer screening,
05:32risk reduction approaches,
05:34cancer surveillance approaches, and, precision
05:37therapies for sure.
05:39Along the way in the
05:40spectrum of care, we know
05:42that with this germline testing
05:44that there are hereditary cancer
05:45implications
05:47to always keep in mind.
05:49And these impacts span not
05:51only from the individual,
05:52but to their families and
05:54to then populations in general,
05:56one family at a time,
05:57but also thinking of some
05:58population level approaches.
06:01So in ASCO twenty twenty
06:02four, the spectrum of studies
06:04for germline testing really spanned
06:06the gamut in terms of
06:08thinking of various different topics.
06:10So the variant spectrum, what
06:12are we learning about,
06:14genetic mutations or pathogenic variants
06:16that are involved in cancers?
06:19What does this mean for
06:20cascade testing in families, and
06:22what are strategies to develop
06:23for cascade testing?
06:25Impact of germline testing for
06:26screening and treatment approaches.
06:29There was some really nice,
06:30abstracts about cancer disparities as
06:32well and,
06:34thinking about this more globally
06:35across populations,
06:37United States, but then across,
06:39really across the world. And
06:41then clinical implementation was a
06:42big area of
06:44work that was being done
06:45and presented at ASCO,
06:47thinking about how to increase
06:49access to germline testing,
06:51and really think about this
06:52in a more broad based
06:53way.
06:55So when we thought about,
06:56the abstracts
06:57to, include in the review
06:59and then decide which ones
07:01to provide for highlights and
07:02updates,
07:03the inclusion for these abstracts
07:06regarding germline testing were those
07:08that,
07:09really kind of, you know,
07:10the main focus of their,
07:12information was about genetic testing
07:13or germline testing as the
07:15subject matter.
07:16This focuses on adult oncology
07:18studies,
07:19and the focus is really
07:21on hereditary cancer genes. There
07:23were some abstracts that were
07:24presented about,
07:26single nucleotide
07:27polymorphisms or SNPs,
07:29some on polygenic risk scores
07:31and things like that. So
07:32we we're not going to
07:33be touching on those, and
07:34we won't be focusing here
07:35at least on the somatic
07:37or genomic,
07:38only types of studies that
07:39are shown here, really keeping
07:40the focus on germline testing.
07:43So overall, this came to
07:44about fifty three abstracts that
07:45we reviewed,
07:47and then selected some for,
07:49highlighting here today.
07:51As far as the, tracked
07:53categories,
07:54from ASCO,
07:56you can see here that
07:57the majority of these fell
07:58under prevention, risk, and genetics,
08:00and these spanned several of
08:02the topics that I just
08:03mentioned already,
08:04in terms of the spectrum
08:05of the work that was
08:06going on. There were some
08:07abstracts, and it was very
08:08interesting from a global perspective
08:10about, genetic testing and germline
08:12variants in Brazil, China, Rwanda.
08:15And there were abstracts in
08:17the veterans populations. So I
08:19really encourage everyone to kind
08:20of go back and think
08:21about, like, maybe just perusing
08:23some of these abstracts when
08:24it comes to germline genetics
08:25and genomics since we won't
08:26be able to cover everything,
08:27obviously, today.
08:29They spanned,
08:31breast cancer,
08:32topics, care delivery and models
08:34of care,
08:35genitourinary
08:36cancers,
08:37gynecologic cancers, lung cancer, which
08:39was extremely interesting,
08:42quality of care health services,
08:44gastrointestinal
08:45cancers, and melanomas.
08:47And some of these topic
08:48areas fell under prevention risk
08:49and genetics, so it's just
08:51a matter of, you know,
08:51kind of thinking through the
08:53different tracks and and finding
08:54those interesting abstracts.
08:57So I'm actually just going
08:58to give highlights on four
08:59abstracts today, for the sake
09:01of time.
09:02These were the oral abstracts
09:04and rapid oral abstracts, that
09:06were presented focused on germline
09:07testing.
09:08So, I'm not gonna read
09:10all of these, but I'm
09:10I picked a few of
09:11these to be able to
09:12present on today since these
09:13were the ones that were
09:14given more attention, at the
09:16ASCO meeting.
09:19So the first abstract we're
09:20gonna, just summarize here is,
09:23abstract one zero five zero
09:24three, which was presented as
09:26an oral presentation
09:27by professor Lambertini from University
09:29of Genova. And this is
09:31about the clinical behavior of
09:32breast cancer in young BRCA
09:34carriers and the prognostic impact
09:36of the timing of genetic
09:37testing pre or post diagnosis,
09:39and this was an international
09:41cohort study.
09:43So the rationale for this
09:44particular study was that, the
09:46clinical behavior of breast cancer
09:48in young, BRCA one and
09:50two carriers isn't fully understood
09:52and deserves some more study.
09:54And then thinking about the
09:56impact of the timing of
09:57genetic testing before or at
09:59the time of diagnosis,
10:00looking at tumor characteristics and
10:02outcomes.
10:03This was a very large
10:04study. It included,
10:06sites from Europe, Asia, North
10:08America, Australia, and Latin America.
10:11The key inclusion criteria were
10:13stages one to three invasive
10:14breast cancer,
10:15diagnosis really over a twenty
10:17year time span. The key
10:19here is this was a
10:20young population of patients. So
10:22really looking at this in
10:23the early onset space,
10:25age, at diagnosis of forty
10:27years or younger
10:28with a known pathogenic variant
10:30or mutation in BRCA one
10:32or BRCA two.
10:34This total the total cohort
10:35for dialysis included, four thousand
10:37seven hundred and fifty two
10:38patients with BRCA one or
10:40two pathogenic variants. The median
10:42follow-up was seven point eight
10:44years as I'm and I
10:44mentioned there was some really
10:46great geographic diversity,
10:47in this study.
10:50So the key findings and,
10:51these actually slides, I wanna
10:53give credit to, Allison Kurian,
10:55doctor Kurian, who was the
10:56discussant for, these, oral abstracts
10:59at ASCO.
11:00And kind of the, highlight
11:02findings here from this particular
11:03study was that compared to
11:05BRCA two carriers,
11:07BRCA one carriers had, were
11:09significantly younger, at the time
11:11of diagnosis,
11:13had more grade three tumors
11:14and triple negative disease. These
11:16are well known features, of
11:17BRCA one,
11:19and, had less nodal involvement
11:21as well as lobular histology.
11:24In looking at second primary
11:26breast cancers and also new
11:28primary non breast cancers, which
11:30can happen with,
11:32BRCA mutation carriers and, carriers
11:34of other gene mutations.
11:36There was this was significantly
11:37more frequent among BRCA one
11:39carriers,
11:41compared to BRCA two carriers.
11:43And BRCA one carriers had
11:44worse disease free survival and
11:46overall survival in the first
11:47five years.
11:49And, what was interesting was
11:50that for BRCA two, there
11:52seems to be this sort
11:53of constant increase in,
11:56recurrence over time, which kind
11:59of shows the different sort
12:00of clinical behavior of,
12:02these two genes.
12:04And then looking at compared
12:05to patients tested at breast
12:06cancer diagnosis, those tested before
12:08breast cancer diagnosis
12:10had smaller tumors and less
12:12nodal involvement
12:13with a trend toward better
12:14disease free and overall survival
12:16when that prediagnostic genetic testing
12:19was implemented.
12:20So the key takeaways from
12:22this particular abstract was that
12:24this was a global study
12:26that, it really looked at
12:27the clinical behavior of breast
12:29cancer, particularly
12:30in a very young population,
12:33of patients diagnosed at age
12:35forty or younger,
12:37and looked at the timing
12:38of that,
12:40genetic testing with their diagnosis,
12:43and thinking about how to
12:44better mutations. And then really
12:46so what, was brought up
12:47as, you know, at the
12:47session as well as when
12:48we think about this
12:49is that there really wasn't
12:51data. Some of the limitations
12:52is that no data on
12:52the,
12:54prediagnosis
13:00screening. Was MRI used? How
13:02was the, cancer picked up,
13:04which could have impacted some
13:05of the outcomes that we're
13:06seeing, in terms of pre
13:08versus at diagnosis for genetic
13:10testing?
13:11And, of course, there was
13:12health care system variability in
13:13this cohort. So how do
13:15we think about this? We
13:16need to factor in some
13:17of that variability when we
13:18interpret the outcomes.
13:21The second abstract that was
13:23presented at the oral session
13:24was, by doctor Shivak from,
13:27Duke University.
13:28This was looking at the
13:29association between rare pathogenic variants
13:32in established cancerous genes and
13:33the diagnosis of single cancers
13:36and multiple cancers in individuals.
13:39And this was a study
13:40from the UK Biobank.
13:43And the rationale for this
13:44study is that, really, as
13:46we've arrived at where we
13:47have in terms of genetic
13:49testing or germline testing, across
13:51the cancer spectrum, most of
13:52the original studies and several
13:54of the follow-up studies have
13:55been family based studies,
13:58or case only studies with
13:59external reference controls, which could
14:02actually have led to some
14:03selection bias. And when we
14:05roll these,
14:06testing practices out in in
14:08practice,
14:09we may not always find
14:11some of the results that
14:12we see in the original
14:13reports and,
14:15that have been published.
14:16So here, they took a
14:17population based approach in looking
14:19at the UK Biobank, looking
14:21at the relationship between these
14:22rare pathogenic variants in ninety
14:24six cancerous genes across eleven
14:27cancer types,
14:28bladder cancer, breast cancer, CNS
14:30cancers, colorectal, lung, melanoma, ovary,
14:33pancreas, prostate, renal, and thyroid
14:36cancers.
14:38So the study design for
14:39this particular study was,
14:41looking at the UK Biobank
14:43whole exome sequencing data.
14:45This really included primarily the
14:47white ray, white race and
14:48ethnicity individuals,
14:50and looking at,
14:52associations with,
14:54the various cancers
14:55as the data was available
14:57in the UK Biobank database.
15:00And then looking at this
15:00from the outcome perspective for
15:02one of those eleven cancers
15:03or multiples in terms of
15:05prevalent cases and controls.
15:09These were the demographics of
15:10the cohort. So you can
15:12see it was a very
15:12large,
15:14cohort looking at the UK
15:15Biobank, over twenty five thousand
15:17cases,
15:18over a hundred and fifty
15:19thousand controls.
15:20And you can see some
15:21of the characteristics here in
15:22terms of median age at
15:24diagnosis,
15:25and those individuals that had
15:26two or more cancer diagnoses
15:28represented about eight percent of
15:30the population.
15:33In looking at the results,
15:35what we can see here
15:36is that the presence of
15:37rare pathogenic variants were associated
15:39with,
15:40various different
15:42levels of cancer,
15:44development.
15:45And it was very interesting
15:47about looking at whether there
15:48was syndromic overlap or beyond
15:51syndromic overlap when looking at
15:52this from a population based
15:54approach.
15:55And so the odds ratio
15:56was, one point eight for
15:58the presence of these rare
15:59pathogenic variants for having,
16:01more
16:02greater than or equal to
16:03one cancer development and was
16:05significantly higher if an individual
16:06had two or more cancers,
16:08that were developed.
16:12So in looking at specific
16:13syndromes and then looking at
16:15specific genes,
16:17first looking at the homologous
16:18recombination repair genes. So these
16:20would be classically linked with
16:22hereditary breast and ovarian cancer
16:23syndrome.
16:24What we see here, the
16:25cancers are, if it's hard
16:26to read, breast cancer, ovarian
16:28cancer, pancreatic cancer, and prostate
16:30cancer,
16:31looking at the odds ratios
16:32and the various genes, in
16:34terms of looking at rare
16:35pathogenic variants.
16:36So BRCA two was associated
16:38with all four of these
16:39genes, in this analysis.
16:42And ATM was also associated
16:43with three of the, cancers,
16:46looking at prostate, breast, and
16:47pancreatic cancers. So,
16:49actually, BRCA two associated with
16:51all four cancers.
16:53When we look at by
16:54cancer type,
16:56if you look at breast
16:57cancer, several of those genes
16:58were associated with breast cancer
17:00in this analysis, ATM, BRCA
17:03one and two, and PALB
17:04two. Ovarian cancer actually had
17:06some of the highest odds
17:07ratios
17:08when looking at the various
17:10genes,
17:11across this spectrum of the
17:13cancer genes tested.
17:15Prostate cancer was associated with
17:16BRCA two, ATM, and check
17:18two, and pancreatic cancer associated
17:20with ATM and BRCA two.
17:24When looking at Lynch syndrome
17:25genes, these were the four
17:27genes that were,
17:29came out with positive findings
17:30in terms of association
17:32with bladder cancer, colorectal cancer,
17:34and ovarian cancer.
17:36EpCAM, one of the genes
17:38also involved in the Lynch
17:39syndrome spectrum, was not associated
17:41actually in this particular analysis
17:42with any of these cancers.
17:46And then I thought this
17:47was actually very interesting when
17:49we take a population based
17:50approach or when we roll
17:51genetic testing out in clinical
17:53practice. What we start to
17:54see is some,
17:56differences or deviations from these
17:58original classic syndromes
18:00in terms of findings of
18:02genetic mutations in patients that
18:03may or may not match
18:04their clinical picture or their
18:05family history. So these were
18:07some of the associations of
18:08genes linked with, various different
18:10cancers that are shown here
18:12in terms of genes linked
18:13with bladder cancer,
18:14CNS cancers, lung cancer, melanoma,
18:17and some of the additional
18:18results that are shown here.
18:20One of the interesting findings
18:21and questions that we get
18:22in practice is about germline
18:24testing for lung cancer patients.
18:26This is quite a frequent
18:27question, and we see this
18:29a lot for patients who
18:30themselves have lung cancer or
18:31have a strong family history.
18:33So, you know, interesting that
18:34some of this data is
18:35pointing to potentially ATM, BRCA
18:38two having higher odds ratios,
18:40of risk for lung cancer.
18:42How we roll this out
18:43in clinical practice is to
18:44be determined, but this gives
18:46us some, you know, information
18:47about thinking about various cancers
18:48where we may not have
18:49strong clinical paradigms quite yet
18:51in terms of thinking of
18:52germline testing.
18:54So some of these associations
18:55were classical with the syndromes
18:57and others were not classical.
18:59And so, again, that just
19:00points to what we see
19:00in clinical practice.
19:02Some of the limitations of
19:03this, study were that this
19:04was looking at prevalent versus
19:06incident cancers, and so there
19:07may be a survival bias
19:08when we think about some
19:09of the results.
19:11As I mentioned, it was
19:12all European ancestry, so this
19:13does need to be validated
19:14across ancestral populations.
19:17And perhaps were variants being
19:18missed since this was whole
19:20exome sequencing data, not full
19:22whole genome sequencing data, but
19:24really kind of thinking about
19:25this and pulling that together.
19:28So the key takeaways from
19:30the,
19:30study team was that these
19:32rare pathogenic variants are associated
19:34with single and multiple cancers,
19:37particularly in DNA repair genes
19:39associated with non canonical cancers,
19:41nonsyndromic,
19:42findings. And these population based
19:44approaches do build on,
19:46existing data largely from these
19:48family and case only studies.
19:50And the discussant, doctor Kurian,
19:52you know, really had a
19:53nice slide that I thought
19:54was interesting to show in
19:55looking at who should receive
19:56genetic testing. She pulled together
19:58a summary of the recommendations
20:00for germline testing from ASCO,
20:02SSO, and NCCN
20:04showing here that this really
20:05does span a significant amount
20:07of the, cancer population.
20:09But does it happen in
20:10practice? She presented some of
20:11her data looking at the
20:12Georgia and California SEER registry
20:14data that showed probably this
20:16is happening at about a
20:17forty to fifty percent rate
20:19even in, populations that have,
20:21say, you know, some clear
20:23indications for germline testing.
20:25And when we think about
20:26putting this in the scope
20:27of then how can this
20:28be applied,
20:29in the real world, can
20:31we test everyone? You know,
20:33factors that affect implementation of
20:35being able to test everyone
20:36are, you know, patient identification
20:38and awareness and practice, insurance
20:40coverage for genetic testing,
20:42limited access to genetic counselors,
20:44and developing a clinical workflow
20:45that'll help
20:47encompass the patients that need
20:48to be seen.
20:51Moving forward, rapid oral abstract
20:52one zero five zero nine
20:54was looking at the frequency
20:56of HER2 low breast cancer
20:58among BRCA one, BRCA two,
21:00and PALB two mutation carriers.
21:02This was presented by doctor
21:03Goldblatt from University of Pennsylvania.
21:06And this was based upon,
21:08the premise for this particular
21:09study was from the results
21:10from the DESTINY BRESTO four
21:12trial,
21:13that was published already in
21:14the New England Journal of
21:15Medicine looking at trastuzumab
21:17daruxedacam
21:18in previously treated HER2 low
21:20advanced breast cancer.
21:22And this particular study qualified,
21:24HER2 low as one plus
21:26IHC
21:26or two plus IHC with,
21:29in situ hybridization
21:30negative.
21:31And these were all patients
21:32that were hormone receptor positive.
21:35And what was, reported and
21:36seen was that the median
21:37progression free survival was superior
21:40for the, trastuzumab daraxdacan group,
21:43compared to physician's choice as
21:45well as overall survival, leading
21:47to FDA approval for this
21:48drug in this, particular for
21:50this particular indication.
21:52Now what's interesting is particularly
21:54for BRCA one,
21:55these cancers are triple negative
21:57breast cancers. So these are
21:59HER2 classically,
22:00characterized as HER2 negative along
22:02with, hormone receptor negative.
22:05But when we look across
22:06different studies, BRCA one,
22:08is, you know, in terms
22:10of rates of saying they're
22:11HER2 negative, maybe two to
22:12ten percent,
22:14of these, patients. They have
22:15a lower frequency of HER2
22:17expression. So two to ten
22:18percent have lower frequency.
22:21BRC one and two associated
22:22breast cancers have lower frequency
22:23of HER2 overexpression.
22:25And so what
22:26this particular study was looking
22:28at is how can we
22:29bring that kind of testing
22:31technology to the space of
22:33BRCA carriers,
22:34and is there any reclassification
22:36of patients that might have
22:38been deemed as HER2 negative
22:40and actually,
22:41thinking that they may be
22:42HER2 low and qualify,
22:44for this, type of treatment.
22:47So this particular study was
22:48looking at identifying the prevalence
22:50of HER2 low breast cancer
22:51among individuals with germline BRCA
22:53one or two or PALB
22:55two, carriers,
22:57and those that had been
22:58classified as HER2 negative. It
23:00was a retrospective
23:01observational study, at the University
23:03of Pennsylvania.
23:05And what they found was
23:06that forty seven percent of
23:07breast cancers that would previously
23:09have been categorized as HER2
23:10negative
23:11would have then been now
23:13reclassified
23:13as HER2 low, So would
23:15have been eligible for trastuzumab,
23:18duroxicam,
23:19in terms of looking at
23:20this new way of, classifying
23:22this. And also interesting was
23:24comparing the metastatic biopsy site
23:26with the corresponding primary site,
23:28twenty five percent were HER2
23:30low at the metastatic site
23:31when the original biopsy site
23:32was classified as HER2 negative.
23:34So potentially thinking of biopsy
23:37metastatic sites if that is,
23:39possible and clinically indicated.
23:42So the takeaways from this
23:43particular abstract was that a
23:45substantial portion of BRCA one
23:47or two, carriers of pathogenic
23:48variants
23:49who have HER2 negative disease
23:51actually could express low levels
23:53of HER2 and to maybe
23:54think about this in practice.
23:57The sample size is very
23:58low for PALB two, so
23:59they couldn't really draw conclusions.
24:01And so it may be,
24:03a a thought in terms
24:04of reevaluation of HER2 status
24:06in metastatic samples as well.
24:09And then the final abstract
24:10was, looking at real world,
24:12cancer care utilization
24:14among breast cancer patients,
24:16with germline variants of uncertain
24:18significance. This was presented by
24:19doctor Allison Kurian from Stanford
24:21University.
24:23And the, rationale for this
24:25particular abstract was that when
24:27we do genetic testing, we
24:29can get three kind of
24:30general types of results back
24:32from genetic testing. One of
24:33which is pathogenic or likely
24:35pathogenic variants or mutations.
24:38The second of which is
24:39variants of uncertain significance.
24:41So these do not impact
24:43clinical management,
24:44but they are in patient
24:45reports. And so the question
24:47is, is there overinterpretation
24:49of these results that informs
24:51management when it's not ready
24:52to inform management?
24:54And then, of course, negative
24:55results.
24:56So the concern in practice
24:57is is our VUS results
24:59prompting overtreatment as there were
25:00some signals in the literature
25:02that there can be recommendations
25:04for,
25:05surgical management or, you know,
25:07other kinds of cancer risk
25:08reduction or screening approaches,
25:10particularly when VUS results return.
25:13So this particular study looked
25:15at linked databases from a
25:16germline testing company
25:18and linking that with insurance
25:20claims.
25:21This was looking at, a
25:23a study sample population of
25:24females with breast cancer,
25:26looking at, those patients that
25:28had had multi gene panel
25:29testing,
25:30and the genes that were
25:32analyzed looked at those genes
25:33that were linked with, high
25:34breast cancer risk as well
25:36as moderate breast cancer risk.
25:40And, really, the summary of
25:42the results is that the
25:43analytic sample included over twelve
25:45thousand,
25:46patients in this particular analysis.
25:49The pathogenic variant rate in
25:50the breast cancer genes was
25:51six point four percent, and
25:53the VUS rate was ten
25:54point four percent.
25:56The most common gene with
25:57a pathogenic variant in this,
25:59group was CHEK two, and
26:00the most common gene with
26:01variants of uncertain significance was
26:03ATM.
26:04And, really, these results showed
26:06that breast cancer patients with
26:07VUSs
26:08were treated like other patients
26:10who tested negative,
26:12which is the way that
26:13it really should roll forward
26:15in practice. No overuse of
26:16bilateral mastectomy,
26:18imaging, PARP inhibitors, or platinum
26:20chemotherapy.
26:21There was a signal, though,
26:22that there potentially could be
26:24more, bilateral stopping of oophorectomies
26:26in these patients,
26:28potentially, as well as selection
26:29of the type of chemotherapeutic
26:31regimens with VUSs versus negative,
26:34which does deserve further study.
26:37And some of the limitations
26:38was that claims data lack
26:40some clinical details, so it
26:41does deserve follow-up, but was
26:43somewhat reassuring when we think
26:45about how VUSs are managed
26:46in practice and that these
26:48educational efforts to providers and
26:49patients
26:50are really important to keep
26:51pursuing.
26:54So as a wrap up,
26:55I did also want to
26:56bring forward a few selected
26:58poster presentations. I'm not gonna
26:59give the detail of these
27:01presentations,
27:02for the sake of time,
27:03but I would point you
27:05to looking up these posters
27:06and their abstracts,
27:08one of which was, neoadjuvant
27:10therapy in patients with pancreatic
27:12ductal adenocarcinoma with germline DNA,
27:14damage repair mutations.
27:16This was a dual institution
27:17retrospective study, and they found
27:20that, for those patients that
27:21had resectable borderline,
27:23or locally advanced, PDAC,
27:26that also carry germline DNA
27:28repair mutations
27:29had significantly prolonged disease free
27:31and overall survival,
27:33with,
27:34improved complete or near complete
27:36pathologic response rates when they
27:38were given neoadjuvant
27:39for farinox.
27:41So, a really exciting area
27:43to think about how genetics
27:44can be implemented,
27:46across cancer types and also
27:47with using standard therapies and,
27:49of course, new and emerging
27:50therapies as well.
27:52There was a study out
27:52of City of Hope that
27:53was the INSPIRE study looking
27:55at more broad based, universal
27:58germline testing for cancer susceptibility.
28:00And this particular study looked
28:02at in person or remote
28:04approaches for consenting and testing.
28:06They approached twenty six thousand
28:07patients and eighty two percent
28:09of those patients enrolled in
28:10the study,
28:11and the, pathogenic variant rate
28:13was twenty percent,
28:14for this particular,
28:16population.
28:17So they're really looking at
28:18a scalable model to increase
28:20germline testing across more unselected
28:22populations,
28:24in their catchment.
28:26Another study was looking at
28:28remote delivery of cancer genetic
28:30testing in veterans with metastatic
28:32prostate cancer. So this was
28:33the million veteran program,
28:35a really great abstract looking
28:37at ways to, again, enhance
28:38access to germline testing and
28:40engage a population where there
28:42is a uniform health care
28:43system.
28:44And so this was, again,
28:45using sort of a remote
28:47outreach approach using informational letters
28:49sent to all eligible participants.
28:51Thirty five percent of the
28:52veterans completed the informed consent
28:54and, twenty three percent of
28:56them
28:57completed testing.
28:58And the the pathogenic variant
29:00rate in this population of
29:01metastatic prostate cancer patients was
29:03thirteen percent. So, again, thinking
29:05of ruling out these into
29:06large integrated health care systems,
29:08for germline testing. And then
29:10implementation of universal germline testing
29:12in underserved patients with young
29:14onset GI cancers.
29:16This was looking at
29:18identifying and then reducing,
29:20barriers and facilitators to germline
29:22testing.
29:22Here, they looked at patients
29:24diagnosed with GI cancers between
29:25the ages of eighteen to
29:26fifty,
29:28and looking at what was
29:29the, ability to engage and
29:31uptake with, genetic counseling and
29:33genetic testing.
29:35The mutation rate here was
29:36seventeen percent for these patients,
29:38but what they did identify
29:40were barriers such as cancer
29:41care delivery,
29:43issues, you know, being,
29:44referrals, uptake of genetic testing,
29:46and timeliness of testing, availability
29:49of genetic counseling as key
29:50barriers to address for germline
29:52testing.
29:54So some of these conclusions,
29:56really, as I mentioned, so
29:57germline testing is now becoming
29:58more and more central in
30:00terms of various aspects of
30:01cancer care, and risk assessment.
30:03Some of the key insights
30:04from ASCO twenty twenty four
30:06were this growing knowledge of
30:08the germline variant spectrum across
30:10cancer,
30:10populations,
30:12The application of germline genetics
30:13for novel therapies into,
30:16thinking about this,
30:17when identifying the germline spectrum
30:19across populations.
30:21Care delivery models,
30:23really do help to enhance
30:24access to germline testing with
30:25demonstrated feasibility,
30:27and there's greater systems based
30:29approaches that are needed across
30:30populations
30:31to expand germline testing.
30:34So I'm just gonna end
30:35with,
30:35you know, a plug for
30:36our cancer genetics and prevention
30:38program. We are growing, and
30:39we are addressing the needs
30:40of our patient populations across
30:42the state of Connecticut.
30:44You know, really looking at
30:45what we learned from ASCO
30:46twenty twenty four, we are
30:48addressing a lot of the
30:49needs for patients who have
30:51cancers, who are at risk
30:52for cancers, and thinking about
30:54this from a disparities and
30:55equity lens as we roll
30:57forth our programs.
30:58These are our physicians in
31:00our program
31:01as well as our, genetic
31:02counselors and clinical genetics coordinators
31:05as part of our fast
31:06track approach.
31:08And these are our, current
31:10signatures of care, that we
31:12are growing to really help
31:13enhance access to germline testing.
31:15The classic genetic counseling model
31:17with referral to genetic counselors.
31:19We've developed a new fast
31:20track program where patients can
31:22receive expedited genetic testing through
31:24our program by coordinators,
31:26and be able to build
31:27in, how we can then
31:28implement, more rapid disclosure of
31:30results to help with their
31:31treatments,
31:33and return of these results
31:34to patients and providers.
31:36And we're going to be,
31:38significantly rolling out point of
31:39care testing
31:40such that, providers can actually
31:42offer genetic testing in their
31:43clinics. This has been piloted
31:45by, physicians over the last
31:47few years, really who are
31:48here today.
31:50And I think this is
31:50a really important model to
31:52roll forth as well as
31:53we know more and more
31:54patients need genetic testing. So,
31:56we're here to help support
31:58the needs of our patients
31:59and our health care providers.
32:01So thank you again for,
32:03this time this morning in
32:04terms of talking about advances
32:06in germline testing. I have
32:07my email here at the
32:08bottom if anyone would like
32:09to reach out to me.
32:10Thank you.
32:13Oh, okay.
32:17Thank you, doctor Gary. We
32:18maybe have time for one
32:19or two questions, and then
32:20we're gonna move to the
32:21GU session. Does anyone have
32:22questions?
32:25No questions.
32:27That's fine.
32:29Well, maybe I'll I'll ask
32:30a question. So,
32:32what are you what are
32:33you most excited about in
32:34terms of really,
32:36sort of further developing the
32:38cancer genetics program? Like, what's
32:40what's the low hanging fruit
32:42that you're really excited to
32:43accomplish this year? Mhmm.
32:45Yeah. I think,
32:47two things. One is really
32:49this ability to enhance
32:58implementation metrics of this. So
32:59be able to have data
33:00that we can present back
33:02to say what works, what
33:03doesn't work, how, how much
33:05you know, where do we
33:06need to improve. So that's
33:07one is an implementation, you
33:08know,
33:10work. And the second is
33:11really cancer disparities.
33:13So building out,
33:15tools,
33:16resources,
33:17and models of care that
33:18really do address,
33:20our populations across the state
33:22of Connecticut. We have populations
33:24that really have,
33:26significant diversity in terms of
33:28racial ethnic diversity, language diversity,
33:32educational, you know,
33:34you know, education along the
33:35spectrum, so of care and,
33:38you know, being able to
33:39understand cancer genetics. And so
33:41we really want to have,
33:42you know, significantly improved outreach
33:45and engagement across diverse populations.
33:48So I'm really excited about
33:49those things. Great. Thank you,
33:51doctor Gary. Thank you.
33:58So we are now moving
33:59into some of our disease
34:01specific content. So I'd like
34:02to invite up doctor Joseph
34:04Kim. He will be giving
34:06the
34:07hot topics in twenty twenty
34:09four GU cancers.
34:10And some of you have
34:11noticed as, doctor Herbst mentioned
34:13this morning, we've reorganized,
34:15the structure of the day.
34:17So these hot topics are
34:18meant to be quick updates
34:20on really key impactful abstracts
34:21that have occurred over the
34:22course of the year at
34:23ASCO and other meetings. So
34:24doctor Kim, please come up.
34:26And then, maybe what I'll
34:27do is introduce and mention
34:29who will be doing the
34:30case discussion, then you can
34:31just turn over to doctor
34:32Petrillac and team. So doctor
34:34Dan Petrillac will be moderating
34:35the case discussions for this
34:37session,
34:38and presenters in that will
34:39be doctors Michael Hurwitz,
34:41Matthew Austin, and Yi An.
34:43Thank you.
35:09Thank thank you. Good morning,
35:10everyone. I'm Joseph Kim. I'm,
35:12one of the geomedical oncologists.
35:15Good to see everyone today.
35:16So my job is to
35:17talk about hot topics in,
35:19geocancers
35:19in two thousand twenty four.
35:20So what I will do
35:21is I will just highlight
35:22some data that was recently
35:23presented at ASMO as well
35:25as, recent ASCO.
35:26And I was trying to
35:27also highlight some research program
35:29that we have in our
35:29geocrT.
35:32So I'm gonna touch on
35:33a couple of trial about
35:34three trials in urethralia carcinoma
35:35and couple of trials in
35:36prostate cancer.
35:38So the first one I'm
35:39gonna dive in the first
35:40one. So this is a
35:41phase three study of, you
35:43know, neoadjuvant dervalumab plus chemotherapy
35:45followed by radical cystectomy
35:47and adjuvant divalumab
35:48in patients with most invasive
35:50bladder cancer.
35:51So we know that the
35:52neoadjuvant chemotherapy improves survival in
35:54patients with a, a mostly
35:55invasive urethral carcinoma, but we
35:57know that about fifty percent
35:58of patients do, experience, recurrence
36:00within the first three years.
36:02We also know that the
36:03immune check inhibitors are used
36:05in adjuvant setting,
36:07and it's been shown to
36:08improve disease free survival in
36:09couple of trials, CheckMate two
36:11seventy four, which led to
36:12the approval of the nivolumab
36:14and ambassador study, which I'm
36:15gonna discuss later, which again
36:17show improvement DFS with, pembrolizumab.
36:20So perioperative immune checkpoint inhibitors
36:22could improve long term clinical
36:24outcomes by priming the,
36:26priming the anti tumor immunity
36:28before the surgery and also
36:30eradicating
36:30the micrometocyte disease after surgery.
36:36Perioperative immune checkpoint inhibitors,
36:38darvelumab combined with a neoadjuvant
36:40chemotherapy in cisplatin eligible patients
36:43with most invasive bladder cancer.
36:46So this was a study
36:47design. They accrued over a
36:48thousand patients. Again, they enrolled
36:50patients who are cisplatin eligible,
36:52most invasive bladder cancer. They
36:53also allowed a node positive
36:55patients as well, and also,
36:57some urethral carcinoma with divergent
36:59differentiation pathology.
37:02It might be the darvelumab
37:04arm versus, control arm. In
37:06patients' darvelumab, they received new
37:08adjuvant darvelumab combined with the
37:09new adjuvant chemotherapy.
37:11They were followed by cystectomy,
37:12and they received, about eight
37:13cycles of darolumab in adjuvant
37:15setting. In control arm, they
37:16received just note of care
37:17with the, gem cyst, and
37:19followed by radical cystectomy.
37:21The study was done actually
37:22before the approval of the
37:23nivolumab, so not many of
37:24not many of these patients
37:25received, subsequent, adjuvant immunotherapy.
37:29The there are two, primary
37:30endpoints,
37:31event free survival and the
37:33pathologic,
37:34complete response.
37:35The primary events were defined
37:37as progressive disease,
37:38that precluded radicle cystectomy
37:41or recurrence after radicle cystectomy
37:42or date of expected surgery
37:44in patients who did not
37:45undergo radial cystectomy. In other
37:46words, they had some reason
37:48not to undergo surgery. This
37:49was, was one of the
37:51primary events and also a
37:52death from any causes.
37:55So, this is a capillary
37:57curve I'm showing, event for
37:58survival. Again, this was a
38:00positive trial. Has a ratio
38:01of zero point six eight,
38:02with a significant p value.
38:04As you see here, two
38:06year event for survival rate
38:07was improved by about, ten
38:09percent range or so.
38:11So this was a positive
38:12trial.
38:14If you look at the
38:15subgroup analysis,
38:16of this primary endpoint against,
38:18there's no particular subgroup that
38:20sort of sends out. Many
38:21of these subgroups do favor
38:22the use of tafalumab.
38:25The other,
38:27primary endpoint was the pathology
38:28complete response.
38:29The two analysis,
38:31basically, they show about ten
38:32percent improvement, in the complete
38:34pathologic response rate with the
38:36dorvailumab.
38:39This was overall survival data.
38:41Again, this was positive. They
38:43showed about seven percent improvement
38:44in overall survival rate about
38:46two years. The hazard ratio
38:47is zero one seven five,
38:48moving twenty five percent reduction
38:50in the risk of death
38:51with us again from p
38:52value.
38:55Again, if you look at
38:55the subgroups,
38:57there's no major subgroup that
38:58sort of sends out. All
38:59of the subgroups,
39:00drive benefit from the perioperative
39:01of the value map.
39:06Look looking at the, AE
39:07summaries, again, the adverse events
39:09were comparable between the two
39:11arms.
39:12Grade three events,
39:14outcome of, death, they were
39:16comparable between the two arms.
39:18And also, a treatment related
39:19grade three events, again, were
39:21comparable between the two arms.
39:23There's no major adverse events
39:25leading to discontinuation
39:27of the neoadjuvant chemotherapy or
39:29foregoing radical cystectomy or delaying,
39:31radical cystectomy.
39:37Looking at the in adjuvant
39:38setting, there were few, you
39:40know, adverse events during, in
39:42the dravolumab as expected,
39:43but there's no major safety
39:45signals or rays, in the
39:46adjuvant use of dravolumab.
39:50So in conclusions,
39:52nigra was the first phase
39:53three perioperative immunotypic study in
39:55most invasive bladder cancer, and,
39:57it led to improvement even
39:58for survival and overall survival.
40:01The benefits were consistent across,
40:03subgroups.
40:05Pathological response rate as well
40:06as, West benefits supports the
40:08period of approach of darvelumab.
40:11And adding dorva to, you
40:12know, neoadjuvant chemotherapy appears to
40:14be tolerable and manageable with
40:15no safety signal.
40:17And neoadjuvant development did not
40:19delay any surgery, will impact
40:21the ability of patients to
40:22undergo radical cystectomy.
40:25So we actually redo it.
40:26There are several trials going
40:27on in preoperative setting. We
40:29have, one of the studies
40:30available in our GOCRT.
40:32That's a Voregut trial. So
40:33just to cover other studies
40:35that are going on, so
40:36patients who are cisplatin eligible,
40:38there's a keynote a six
40:39six, also a CheckMate two
40:41seventy
40:42six trial looking at their
40:43own IEO combined with the
40:45chemotherapy.
40:46And, also, there's EV based
40:47preoperative trials,
40:49using, EV pembrolizumab
40:51in keynote three zero four
40:52trial. This is for patients
40:54with sput ineligible.
40:55And since ineligible patients,
40:57they are comparing EV up,
40:59up, with, you know, upfront
41:01cystectomy or, pembrolizumab as a
41:03monotherapy.
41:04We have a VORGA trial
41:05sponsored by AstraZeneca.
41:07This is for patients with
41:08the cisplatin ineligible.
41:09This is with EV plus
41:11durvalumab with or without the
41:12trimilumab, which is CTLA for
41:14antibody versus upfront radical cystectomy.
41:17Again, there's an adjuvant IO,
41:19given
41:20in the, experimental groups.
41:23The next study is an
41:25alliance,
41:26a, o three one one
41:27five zero four, the ambassador
41:29trial. This is a study
41:30looking at the the atrium
41:31pembrolizumab
41:31in patients with most invasive
41:33bladder cancer versus observation
41:35in patients who had radical
41:36cystectomy,
41:37the, update of survival data
41:39here. So,
41:41so, again, this was for
41:42patients with, you know, most
41:43invasive bladder cancer after cystectomy.
41:45If they have a high
41:46risk features defined by y
41:48p t two or greater
41:49were a no positive disease
41:51or positive surgical margin, or
41:53for patients who did not
41:54receive neoadjuvant chemotherapy,
41:56p t three or greater
41:57were no positive patients. They
41:59will randomize receive either pembrolizumab,
42:01which is observation.
42:02Pembrolizumab was given, over one
42:04year. And, again, two, primary
42:07endpoints, disease free survival and
42:08overall survival.
42:10This was a positive trial.
42:12Our hazard ratio is zero
42:13point seven three with a
42:14p value zero point zero
42:15zero two seven,
42:18and the median,
42:20disease free survival from pembrolizumab
42:22was twenty nine point six
42:23months versus fourteen point two
42:25months in the observation arm.
42:28They also looked at different
42:29subgroups by PDL one status.
42:31Again, the PDL one status
42:32appears to be a prognostic
42:34marker, if you are PDL
42:36one positive with defined by
42:37CPS
42:38score, greater than ten. They
42:40have somewhat better, you know,
42:41disease for survival, but, PDL
42:43one status was not a
42:44predictive of, benefit of pembrolizumab.
42:49They also looked at,
42:51by nodal staging as well.
42:52Again, we know that nodal
42:53staging is prognostic, but, again,
42:55the benefit of pembrolizumab was
42:57not limited to a certain
42:58nodal stage. There's a benefits
43:00we're seeing throughout,
43:01different nodal stagings here.
43:04There were some common sites
43:05of metastatic disease and recurrence.
43:08Again, lymph metastatic is very
43:10common in the pelvic lymph
43:11node. Bone metastases,
43:12liver, and lung were common
43:14sites of metastatic disease recurrence.
43:18So at adjuvant pembrolizumab,
43:19it's, again, improved, disease free
43:21survival compared to observation in
43:23patients with high risk muscle
43:24invasive rheumatoid carcinoma after the
43:26radical cystectomy. And the subgroup
43:28analysis show, DFS benefit of
43:30pembrolizumab,
43:31regardless of the PD-one status
43:33and the lymph node status.
43:35Metastatic recurrence were common in
43:37patients on observation group,
43:39and the size for lymph
43:40node metastasis,
43:41lung, bone, and liver.
43:43Based on the, you know,
43:45doubling of the median DFS
43:46and manageable toxicity, the study
43:48supports the use of, pembrolizumab
43:50as one of the therapeutic
43:51options. So this is currently
43:53under review at the FDA.
43:57So I'm gonna switch my
43:58gear to a different, interesting
44:00study called,
44:01Tambola study. This was done
44:03in Denmark.
44:04This was actually a national
44:06nonrandomized,
44:07study,
44:08looking at the c use
44:09of ctDNA as one of
44:11the biomarkers to intervene with
44:13the adjuvant,
44:14tezolizumab. So patients who underwent
44:16radical cystectomy, they were followed
44:18with a, the ctDNA testing.
44:21And when they become positive,
44:23if in other words, if
44:24they have a ctDNA detectable,
44:26they were treated with, tezolizumab,
44:28and they were followed for
44:29up to five years. So,
44:30again, this was not a
44:31randomized controlled trial. I'm looking
44:33at the ctDNA as one
44:34of the biomarkers of intervention.
44:36So,
44:37interestingly,
44:38ctDNA
44:39positivity was not was truly
44:41independent of their clinical high
44:43risk features. About half of
44:44the patients who had no
44:46high risk features on pathology,
44:47they were still had, they
44:49still had, positive ctDNAs.
44:54As you see here, about
44:55fifty six percent of the
44:56patients,
44:57has positive ctDNA following their
44:59radical cystectomy,
45:00and about three quarters of
45:02the patients,
45:02had detectable ctDNA within the
45:05first four month after a
45:06cystectomy.
45:07And the, the median lead
45:09time
45:10over the metastasis identified as
45:12CT scan was about forty
45:13three days.
45:14And of the patients with
45:16ctDNA negative, only two patients
45:17end up developing metastasis on
45:19the CT scan during the
45:21follow-up time.
45:23CtDNA,
45:24positivity
45:25is a prognostic factor for
45:27patients,
45:27who has, you know, positive
45:29ctDNA. They have somewhat worse,
45:31regulatory survival and overall survival.
45:35So based on this data,
45:37there was couple of trials
45:38going on,
45:40InVigor,
45:41o eleven study and also
45:42a modern trial where they're
45:43looking at the ctDNA positivity
45:45as a marker to intervene.
45:47So what they are doing
45:48in these studies is that,
45:50after cystectomy, they follow,
45:52and if they become, c
45:54tDNA positive,
45:55so on this InVigor study,
45:56the patients re randomized receive
45:58either one year of tesolizumab
46:00versus placebo.
46:01If they were c tDNA
46:02negative, again, they'll be followed.
46:04If they become CTN positive,
46:05then they can be intervened
46:06with the adjuvant,
46:07therapy.
46:09Again, this is a modern
46:10trial,
46:12coordinated by the alliance study
46:13group. We we are participating
46:15in this study. Again, the
46:16very similar design to what
46:17I told you with the
46:18InVigor study. Again, here we
46:19are using our AtriVent and
46:21nivolumab with or without riladamab.
46:23Again, this is for patients
46:24with a CTDNA positive, but
46:26for CTDNA negative patient, they'll
46:27be randomized with their, nivolumab
46:30for one year versus
46:31surveillance. Again, if as they
46:33have been followed, if they
46:34become positive,
46:35they'll be randomized they'll be
46:37cross over to, one year
46:38of nivolumab.
46:42So I'm gonna switch my
46:43gear to, the prostate cancer.
46:45So AR degrader is really
46:47making a big, you know,
46:49shift, in prostate cancer. So
46:51BMS, three sixty five is
46:53an orally,
46:54bioavailable
46:55first in class hetero bifunctional
46:57molecule with a two or,
46:58MOA
46:59that degrades and also inhibits
47:01the Andrews receptor as a
47:03heterodimer,
47:04as a, bifunctional
47:05molecule. It has a two
47:06binding moieties, one AR binding
47:08moiety, and then the other
47:09is a CRBN binding moiety.
47:11Idea is to bring Andrews
47:13receptor to a CRBN, which
47:15will facilitate a targeted ubiquination
47:17and subsequent proteasome,
47:18AR degradation.
47:20It also has a AR,
47:21binding moiety and competitively inhibits
47:24the angerson receptor binding.
47:27So this was a phase
47:28one study.
47:29They presented a, the part
47:31b, which is the dose
47:32expansion cohort. They,
47:35tested about four different dose
47:36levels. They were sort of
47:37going into a twice a
47:38day dosing level, starting from
47:40four hundred milligram to a
47:41nine hundred milligram. Again, this
47:42was a patient who actually
47:43fairly, heavily pretreated.
47:47And this is a baseline
47:48characteristics.
47:50Nearly over nearly about fifty
47:51percent of the patients receive
47:52prior chemotherapy
47:54and the both lines of,
47:56RCARP
47:57inhibitors.
48:00So in terms of safety,
48:02there were mention of, prolonged
48:04QTCs.
48:05Some of them are grade
48:06three QTC prolongation. They were
48:08asymptomatic.
48:09About six of about six
48:11patients had QTC prolongations,
48:13on this study,
48:15and, and about four percent
48:17of the patients, had this
48:18adverse event at nine hundred
48:20milligram, twice a day dosing.
48:23There's no treatment rate at,
48:24adverse event that led to
48:25the discontinuation of the study
48:27drug here.
48:29This is a PSA response
48:30by prior treatment and, by
48:32the dosing. It appears to
48:33be a a dose dependent
48:35response,
48:36more responses than nine hundred
48:37milligram dosing, but we have
48:38to keep in mind of
48:39the toxicity that were noted
48:40at the dose level.
48:43This is the PSA response.
48:44And, also, I mean, looking
48:45at the median,
48:46PFS in patients,
48:49based on the AR ligand
48:50bind domain domain mutation status,
48:52the response was seen in
48:54both the wild type as
48:55well as the mutant group.
48:57Again, numbers are quite small,
48:58for a ligand binding domain,
49:01group here, only about eleven
49:02patients. But, again, it seems
49:03to be a good activity,
49:05with this agent regardless of
49:06the AR mutation ligand binding
49:08domain mutation status.
49:11So in summary,
49:13PMS,
49:14nine eight six three sixty
49:15five was in a fairly
49:16well tolerated with manageable safety
49:18profile. There was a concern
49:19some concern for QGC prolongations.
49:23This was they are still
49:23working on the dose optimization
49:25at this time.
49:26They were seeing some antitum
49:28activity in heavily protruded patients
49:29with m cRPC,
49:31all progressing on at least
49:32one line of RP therapy.
49:35This is one of the
49:36first in class two AR
49:37degrader and antagonist with the
49:39potential of overcoming the RP
49:41resistant mCRPC
49:42regardless of their ligand brangulum
49:44main mutation status. We are
49:45in process of having the
49:46study in our, GU CRT.
49:49We do have some other,
49:50AR degrader trials in our
49:52GU CRT. We've been heavily
49:53involved with the studies with
49:54ARVIN seven six six as
49:56well as we have another
49:56AR degraded,
49:58using, from Genentech Roche as
50:00well.
50:01So this is another study.
50:02This was actually one of
50:03the present show presentation
50:05at the ASMA meeting. This
50:06is a piece three trial.
50:08This is basically looking at
50:09the combination of enzalutamide
50:11with radium two to three
50:13versus enzalutamide monotherapy
50:14in patients with, mCRPC
50:16who are asymptomatic
50:18or asymptomatic with, bone metastasis.
50:21So this was the,
50:23again, credits to doctor Fezzasi
50:25who discussed this study.
50:27Really, at the time of
50:28study design, we knew that
50:29enzalutamide
50:30improved, you know, in overall
50:32survival in both pre chemotherapy
50:33setting, post chemotherapy setting, or
50:35from trial, prevarial trial. We
50:37knew that. We also knew
50:38that the radium two three
50:39improved survival based on this,
50:41Assimka trial. This is over
50:42a decade ago,
50:44and there's no overliving toxicities
50:46of these two agents. And
50:47based on the our experience
50:49with combining ADT with the
50:50radiotherapy,
50:51in that perfect sense to
50:52combine these two agents.
50:54We also knew that, born
50:55protective agents, is important in
50:57managing, our patients. It does
50:59prevent the SRE.
51:01We use either a danosinib
51:03or soloduronic
51:04acid, for our patients.
51:06So in terms of the
51:07efficacy and safety data, again,
51:09the patients who randomized to
51:10receive either combo versus enzalutamide
51:12monotherapy.
51:13In met, the primary endpoint
51:14of a radiographic progression for
51:15survival,
51:16difference was about three month.
51:20In terms of the safety,
51:22we have to remember the
51:24earlier trial,
51:25with the ERAS two to
51:27three trial where they used
51:28the radium two to three
51:29with abiraterone.
51:31There was increased fractures, in
51:33patients not receiving point protective
51:35agents. So, actually, they terminate
51:37this, study. And, on this
51:39piece three study, they mandated
51:41the use of the point
51:42protective agent, and it does
51:43have an impact in preventing
51:45the the risk of fracture.
51:50Great. And in terms of
51:52the safety of this study,
51:54overall safety,
51:56AEs were fairly well manageable.
51:57There was an increased incidence
51:59of hypertension
52:00in patients receiving enzalutamide. More
52:02than thirty percent of patients
52:03experience, hypertension. So we have
52:05to close in mind those,
52:06side effects.
52:08And, also,
52:10we have to keep in
52:11mind of the osteonecrosis of
52:12the jaw, which can be
52:13caused by the bone protective
52:15agents. They have not reported
52:16this data, here in this
52:18presentation.
52:20So,
52:21is it practice changing? You
52:23know, we do know that
52:24that, you know, the combination
52:25have an impact on the
52:26RPFS. We like to have
52:28another meaningful clinical endpoint.
52:30In this study, they did
52:32not, show improvement in time
52:34to pain progression or time
52:35to schedule related events.
52:38Looking at the survival data,
52:41there is a difference in
52:42survival,
52:43by about seven month or
52:45so. But if you look
52:46at the first eight month
52:47or so, there's a overlap,
52:49of the,
52:50survivor curve.
52:51We do need to wait
52:53for further statistical
52:54analysis to make sure that,
52:56this also statistical significance. But
52:58clinically speaking, there seems to
52:59be a big difference in
53:00the overall survival
53:03here.
53:04Again, the landscape is changing
53:06constantly. It's not the only
53:07study. Looking at the radium
53:09combinations,
53:10we do have a study
53:11that's led by doctor Michael
53:12Horwis, here at Yale.
53:14We have a Dura trial
53:15combining dose of test chemotherapy
53:17with the radium two to
53:17three. Again, this is a
53:18randomized controlled trial, and there
53:21are also other trials going
53:22on to combining radium two
53:24to three with other, RP
53:25and the chemotherapy and other
53:27setting.
53:29Alright. So,
53:31combining enzalutamide with radium two
53:32three, it does improve,
53:34RPFS
53:35as well as overall survival
53:37in patients with mCRPC,
53:38somewhat earlier setting. We use
53:40radium two three rather in
53:42later stage setting in patients
53:43with a quite significant,
53:44disease symptoms, but this data
53:47support that we may be
53:48able to use this, agent
53:49early in the setting. Again,
53:51the bone protective agent is
53:52mandatory,
53:53you know, for our patient
53:54population, especially when they're receiving
53:55the radium two three.
53:57And the baseline symptoms was
53:59not needed to benefit from
54:01this combination approach.
54:03And the blood pressure management
54:04is very critical,
54:05in patients receiving the, enzalutamide
54:07based therapy.
54:09More data is needed, from
54:11the trial looking at the
54:12ONJ, quality of life, and
54:14also a long term survival
54:16data with additional statistical analysis.
54:20And, with more radium
54:23two three based trials, we'll
54:25be able to optimally use
54:26this agent, for our patients.
54:28And I discussed this, the
54:30data for lutetium one seven
54:31seven. There's pretty emerging,
54:34data, using lutetium one seven
54:36seven as well as other,
54:37radioligand
54:38therapy in prostate cancer.
54:40So, just to kinda summarize,
54:43for urethral carcinoma and the
54:44landscape is changing, we are
54:45using this IO in perioperative
54:47setting and also in adjuvant
54:49setting.
54:50We need to do better.
54:52We are sort of looking
54:53at the use of c
54:54tDNA in identifying patients who
54:55are at risk of, relapse,
54:58of their disease. We're identifying
55:00what they call minimal distress
55:02diseases,
55:03where we can, adequately use
55:04the, the immunotherapy in adjuvant
55:06setting. Also, in prostate cancer,
55:08again,
55:09now AR degrader,
55:11is one of the, agents
55:13of primacyn activity.
55:14AR remains to be a
55:16valid treatment target for, treatment
55:17refractory mCRPC. There's novel targets
55:20such as PSMA and STIP
55:21one.
55:23We have several trials,
55:24into
55:25our portfolio,
55:27though I did not discuss
55:28in MCSPC
55:29setting for newly diagnosed metastatic
55:31cancer since prostate cancer. We
55:32have a PARP one selective
55:33inhibitors as well as a
55:35SIM cap trial looking at
55:36the cytoreductive prostatectomy for our
55:37patients.
55:38We also have a trial
55:40with a Securysa t with
55:41a bipolar androgen therapy,
55:43for, in earlier setting of
55:45mCRPC.
55:46We have several,
55:47bispecific,
55:48trials using CYP one, CD
55:50three bispecific, PSMA, CD three
55:52bispecifics.
55:53These trials
55:54are really ongoing right now.
55:56We have other novel targets
55:58such as the Jira antagonist
55:59and others as well. In
56:00urethral carcinoma, as mentioned, we
56:02have a VORGA trial in
56:04setting using,
56:05EV based,
56:07therapy in adjuvant setting, the
56:09modern trial. Again, looking at
56:11the,
56:12nivolumab with rivolumab,
56:15and we are using our
56:15c tDNA to identify these
56:17patients.
56:18And for frontline setting, we
56:19have a different adenctin for,
56:21antibody conjugate. It can combine
56:22with pembrolizumab
56:23in both frontline setting and
56:25in second line setting as
56:26well. We have a couple
56:27of the STRUCT trials,
56:29using a,
56:30arabulance
56:31combined with gemcitabine.
56:32Doctor. Matt Austin is leading
56:34the study, from our care
56:35centers. We have other novel
56:36targets, in urethral carcinoma.
56:40Very good. Thank you. Any
56:42questions?
56:45No.
56:47I think if not, then
56:48then we'll come on over,
56:49and, we'll moderate the next
56:51session. Thank you.
56:56So good morning.
56:58Great seeing everybody here today.
56:59We're going to be talking
57:00about a couple of cases.
57:02We do use doctor Matt
57:03Austin,
57:04doctor Yi Ahn, and doctor
57:05Michael Hurwitz,
57:06and doctor Cho Kim will
57:07be discussing some of our
57:08kids this morning.
57:14So one of the things
57:15I think that's becoming very
57:16apparent is that inflodimab vedotin
57:19is a critical drug in
57:20the treatment of metastatic urothelial
57:23carcinoma.
57:24And, we presented some data
57:25to ASCO this year,
57:27retrospectively
57:28looking at three trials where
57:29we found
57:30that dose intensity within the
57:32first couple of cycles was
57:33really critical,
57:34to the overall response rate.
57:36So, Mike, you wanna start
57:38talking about this case?
57:40Sure.
57:41Sure.
57:42Can y'all hear me? Yeah.
57:44Okay. Alright. So, this is
57:46a sixty four year old
57:47guy with metastatic
57:48TCC.
57:49He got nirilizumab two hundred,
57:53but not on.
57:55Now it's on. Pretty loud
57:56anyway.
57:58Alright. And he got two
57:59doses of of EV,
58:01which as Dan pointed out,
58:03you know, this dose intensity
58:03really matters quite a bit.
58:05So eight days after his
58:06second dose, he developed generalized
58:07malgias
58:08and a generalized,
58:09exanthemumus
58:11like eruption.
58:12Rashid proddic, it had a
58:13burning sensation, it didn't improve
58:15with diphenhydramine.
58:16And on exam, he had
58:18diffused two to six,
58:19millimeter pink blanchable non tender
58:21macules and thin papules, which
58:22I think we have here.
58:24Right?
58:26That'll be that'll be the
58:27next Oh, that'll be the
58:28next one. Yeah. So so,
58:30you know, what are the
58:31options for this guy? Oh,
58:33yeah. Well, there he is.
58:34Okay.
58:35K.
58:36So
58:37coming back.
58:39And what are the options
58:40for him? And and and
58:41I guess, you know, one
58:42question is, what is the
58:43etiology? Is it definitively
58:45enfortumab? But,
58:46so one, you could treat
58:47through potentially. Another, you could
58:49dose reduce,
58:51from one point two five
58:52by twenty percent, which is
58:54sort of the standard dose
58:55reduction we do with this
58:56agent.
58:57You could hold it,
58:59and start topical steroids,
59:02or you could continue it
59:03and start topical steroids, or
59:04finally, you can continue it
59:05and start oral steroids.
59:08So,
59:10anyway, we can should we
59:12just go through or Sure.
59:12What do people wanna talk
59:13about? We go through that
59:14and,
59:15well, actually yeah. Why don't
59:17we just simply go through
59:18this? This is our rash.
59:19And and these, Mike, these
59:20are the skin manifestations that
59:22we've seen with with, enforidumab.
59:25It can range from anything
59:26from conjunctivitis, maculopapid rash, truncal
59:28rashes,
59:29blisters, dry skin. Hyper hyperpigmentation
59:32is actually something we've seen
59:33in people who've been on
59:34this for a long period
59:35of time. Not really sure
59:36what the etiology of this
59:37is. But the important thing
59:39to keep in mind with
59:40enfortumab
59:42is managing toxicity because you're
59:44looking at the long term,
59:47treatment gained from this particular
59:49drug. So we presented some
59:50data that at ASCO that
59:52you can actually interrupt doses
59:53safely
59:54after the first two cycles
59:55of treatment
59:56without compromising the overall response
59:58rate. So in this particular
01:00:00case, the rash is something
01:00:01that we found from that
01:00:02analysis that generally occurs within
01:00:05the first eight to twelve
01:00:06weeks of treatment. Neuropathy occurs
01:00:08much, much later, and that
01:00:09really you do need to
01:00:10dose reduce or or hold
01:00:11doses at that particular time.
01:00:12But these are the classic
01:00:14rashes that we see in
01:00:15this situation.
01:00:16And,
01:00:18there's an algorithm we use,
01:00:20to, to to, basically treat
01:00:23these particular patients. You know,
01:00:24a grade one rash supportive
01:00:26care is clinically indicated. A
01:00:28grade two rash,
01:00:29you would consider anti infectives
01:00:31as well as topical steroids.
01:00:33And then grade three, you
01:00:34would interrupt
01:00:35and then potentially give oral
01:00:36steroids as well in this
01:00:37particular occasion. The thing you
01:00:39don't wanna see happening is
01:00:40Stevens Johnson syndrome. We've had
01:00:42one case of Stevens Johnson
01:00:44syndrome on the phase one
01:00:45trial. So that's something that's
01:00:47important. In fact, in France,
01:00:49when the drug was first
01:00:50being rolled out,
01:00:51they actually stopped the rollout
01:00:53because of that particular reason.
01:00:54They're patients who were treated
01:00:56through the rash, and they
01:00:57developed grade threes, and they
01:00:58were never put on, topical
01:01:00steroids.
01:01:01Lucinda, if you wanna go
01:01:02back for one second. I
01:01:03think one of the other
01:01:03things that is notable, it
01:01:04doesn't happen that often, obviously,
01:01:06but this at the top
01:01:07here with, like, the red
01:01:08sign there, those are the
01:01:09things you have to worry
01:01:10about. If people are really
01:01:11having pain,
01:01:12if they're febrile, if they're
01:01:14having systemic symptoms, then you
01:01:15really worry that it's one
01:01:16of these really terrible people,
01:01:17Stevens Johnson, etcetera.
01:01:19Exactly. So that's that's, I
01:01:20think, the critical thing. Watch
01:01:22out for Stevens Johnson syndrome.
01:01:24And the neuropathy is is
01:01:25something that we dose reduce
01:01:26or hold doses at at
01:01:27a particular time. That is
01:01:29not as much an influence
01:01:31on the overall response rate.
01:01:32In fact, you can interrupt
01:01:33patients for two and three
01:01:34months, and their disease will
01:01:35remain stable. So I think
01:01:37that that's really a a
01:01:38critical thing towards, patient management.
01:01:41Okay. Number two.
01:01:43Yeah. Thank you. So, case
01:01:44two, forty eight year old
01:01:45male,
01:01:46presented with progressive lower urinary
01:01:48tract symptoms, was noted to
01:01:49have an elevated PSA,
01:01:51which prompted a prostate MRI.
01:01:53He had a PI RADS
01:01:54five lesion in the right
01:01:55side of his prostate
01:01:57and was noted to have
01:01:58this diffuse abnormal signal throughout.
01:02:01Additionally, it was noted to
01:02:01have pathologic appearing lymph nodes
01:02:03in the pelvis.
01:02:05Through urology, given some urinary
01:02:07retention, underwent a TURP,
01:02:09was noted to have a
01:02:10frankly malignant mass.
01:02:12And pathology was a little
01:02:14bit of a curve ball.
01:02:14It showed a mixed high
01:02:16grade neuroendocrine
01:02:17carcinoma
01:02:18combined with the Gleason five
01:02:20plus four prostatic
01:02:22adenocarcinoma.
01:02:24The neuroendocrine component stained classically
01:02:27for neuroendocrine tumor was synaptophysin
01:02:30chromogranin
01:02:30positive with a very, very
01:02:32high k I sixty seven.
01:02:34The adenocarcinoma
01:02:35was negative for neuroendocrine markers
01:02:37and positive for NKX
01:02:39three point one. Patient underwent
01:02:41a PET scan, had widespread
01:02:42diffusely FDG avid. This was
01:02:42not a PSMA PET scan.
01:02:44Lymph nodes in the pelvis,
01:02:44innumerable FDG avid pulmonary nodules,
01:02:45and then markedly hypermetabolic
01:02:47lesions in his liver as
01:02:48well.
01:02:56So quick question, Matt. What
01:02:58was the patient's PSA initial
01:02:59presentation?
01:03:00Relatively low less than twenty.
01:03:01I believe it was probably
01:03:02twelve or thirteen. So this
01:03:04is actually classic for this
01:03:05type of presentation, low PSA.
01:03:08Small cell or neuroendocrine
01:03:10generally does not make
01:03:12a large
01:03:13portion of PSA. In fact,
01:03:15often coexist as you're seeing
01:03:16in this particular case along
01:03:18with the standard adenocarcinoma
01:03:20of the prostate.
01:03:21You make an important point
01:03:22about the PET scan. This
01:03:23patient had an FDG PET.
01:03:26And one of the things
01:03:27that we have noticed has
01:03:28been noticed with pets with
01:03:29PSMA PET scanning is that
01:03:31these are generally cold
01:03:33in the neuroendocrine tumors or
01:03:34the high grade prostate cancers
01:03:36in the liver. We've had
01:03:37some cases where you've actually
01:03:39image positive for small cell,
01:03:41with, with PSMA, but they're
01:03:43very, very rare. But the
01:03:45the one thing you have
01:03:46to be very, very careful
01:03:47of if you use a
01:03:48PSMA PET scan is you
01:03:50can miss hepatic metastases.
01:03:52And, and the other area
01:03:54too that we tend to
01:03:55see false positives in is
01:03:57in the, in the ribs.
01:03:59So what are your options
01:04:00for this patient?
01:04:02Yeah. So I put up
01:04:03four options,
01:04:05and I think it highlights
01:04:06a lot of practice variability
01:04:07in the community.
01:04:09Option one would be triplet
01:04:10therapy for high volume metastatic
01:04:12castrate sensitive prostate cancer,
01:04:14second generation antiandrogen plus hormonal
01:04:16therapy plus taxatyr.
01:04:18Option two, I put doublet
01:04:20chemotherapy with carboplatin docetaxel.
01:04:23Option three would be carboplatinetoposide.
01:04:26Then option four is really
01:04:27extrapolating from the lung cancer
01:04:29literature
01:04:30with small cell lung cancer
01:04:31where we saw,
01:04:33survival benefit with the incorporation
01:04:35of immunotherapy.
01:04:36And I would say in
01:04:37the community and academic centers,
01:04:39there's a lot of variability
01:04:40about the use of immunotherapy
01:04:42and extrapulmonary
01:04:44high grade neuroendocrine carcinoma.
01:04:46So I put that in
01:04:47as a fourth option as
01:04:48well. So this is a
01:04:50castration
01:04:51sensitive patient. Correct? So he's
01:04:52not new diagnosis. New diagnosis.
01:04:54He's not had any treatment
01:04:56pre previously.
01:04:57So, Joe, what do you
01:04:58think? So I think, you
01:05:00know, we have to treat
01:05:01both disease. I mean, the
01:05:02one's gonna kill him will
01:05:03be, you know, neuroendocrine
01:05:04carcinoma with the process, so
01:05:05they'll be the priority. So,
01:05:07actually, I will go with
01:05:08option number three. Again, we
01:05:09have to kinda think about
01:05:10some study that we may,
01:05:11discuss later on, but, carbutoposide
01:05:14is sort of what we
01:05:14have in mind. And I
01:05:15think, you know, given it's
01:05:17a band of carcinoma component,
01:05:18I would definitely start some
01:05:19ADT with the second generation
01:05:21of the antagonist as well.
01:05:22Mike, you agree?
01:05:24Yeah. I think for the
01:05:25most part, I think that,
01:05:26we totally agree with Joe
01:05:27that that the thing that's
01:05:28gonna kill this man is
01:05:30gonna be the neuroendocrine component
01:05:31even though this is a
01:05:32very high grade adeno that's
01:05:34not gonna be a big
01:05:35issue for him comparatively.
01:05:36I think,
01:05:37you know, was it it
01:05:38wasn't definitively small cell. Right?
01:05:41So Nope. It was specifically
01:05:43high grade neuroendocrine carcinoma with
01:05:44high k six. So I
01:05:45I think that, you know,
01:05:46there's a a similar but
01:05:47not identical approach,
01:05:49really primarily driven by the
01:05:51work of Anna Parisio and
01:05:52at at at, MD Anderson
01:05:54of using taxane,
01:05:56platinum combinations. So I think
01:05:58number two would be a
01:05:59reasonable option as well,
01:06:01or even cabazepat you know,
01:06:02you could use cabazepat instead
01:06:03of docetaxel. But,
01:06:05yeah. So I think either
01:06:06two or three would be
01:06:08reasonable. I just want to
01:06:08point out one thing, which
01:06:09is that the only time
01:06:11I ever see this combination,
01:06:12which we see relatively frequently
01:06:14now, is,
01:06:16de novo is when you
01:06:17have Gleason four plus five,
01:06:18five plus four, or five
01:06:19plus five plus five disease.
01:06:20I've never seen it in
01:06:21anything less than that. Right.
01:06:23And so I think or
01:06:24what we now call WHO
01:06:25five disease. In those patients,
01:06:26I'm particularly worried. So if
01:06:28I see anyone with WHO
01:06:29five disease, and if they
01:06:30don't have this, I scan
01:06:32them more frequently because sometimes
01:06:33you'll have the PSA drop
01:06:34to nothing. And then the
01:06:35next thing you know, they've
01:06:36got liver remittance or something
01:06:37else. Exactly. So there there
01:06:38is a difference between de
01:06:40novo versus
01:06:42treatment related. In fact, if
01:06:44you withdraw androgen from
01:06:46prostate cancer cell lines, you'll
01:06:48find that they do develop
01:06:49along the neuroendocrine pathway.
01:06:50So anytime you have a
01:06:51patient who has a low
01:06:52PSA or no PSA,
01:06:54and they start developing systemic
01:06:56symptoms, that's that's a justification
01:06:57for reimaging that patient. Again,
01:06:59preferably with conventional imaging rather
01:07:01than with PSMA PET scans
01:07:02because you may miss hepatic
01:07:03disease. Doctor Ahn, any role
01:07:05for local therapy here?
01:07:07In this situation,
01:07:09really, local therapy with radiation
01:07:10reserved for palliation.
01:07:12The one exception that we
01:07:13will say is that if
01:07:14it's really all go metastatic,
01:07:16we can attempt
01:07:17a reasonably safe local therapy
01:07:19and see if that can
01:07:20delay the need for adjusting
01:07:22the systemic systemic therapy, but
01:07:24it's really case by case.
01:07:25Terrific.
01:07:27Okay. So, Matt, why don't
01:07:28you describe, SWOG two twenty
01:07:30twelve? Yeah. So this is
01:07:31a cooperative group study, and,
01:07:34my clinical practice is in
01:07:35the community. And when I
01:07:36I get excited about with
01:07:37this trial, it's available
01:07:38in probably the majority of
01:07:40the Yale Cancer Center community
01:07:41sites.
01:07:42So it's extrapolating from the
01:07:44small cell lung cancer literature
01:07:45and really trying to answer
01:07:46the question if incorporation of
01:07:48immunotherapy
01:07:49for poorly differentiated
01:07:51high grade
01:07:52extrapulmonary
01:07:53neuroendocrine carcinoma
01:07:55makes a difference. So it's
01:07:56a three arm trial. Arm
01:07:57one is the control arm.
01:07:58They get randomized or in
01:08:00this slide, it's arm three.
01:08:01They get randomized to platinum
01:08:03standard of care chemotherapy with
01:08:05etoposide.
01:08:06And then in the the
01:08:07diagram, arm one is atezolizumab
01:08:09in the induction
01:08:11followed by maintenance atezolizumab.
01:08:13And then arm two is
01:08:14trying to answer a question
01:08:14of whether we really need
01:08:16to continue atezolizumab in the
01:08:18maintenance arm, or is that
01:08:20really only important in the
01:08:22induction setting? And continuing in
01:08:24the maintenance arm may provide
01:08:25additional autoimmune toxicity,
01:08:27but no clinical benefits. So
01:08:29it's a three arm trial
01:08:31and essentially it's trying to
01:08:32answer the question, extrapolating from
01:08:34lung cancer small cell,
01:08:36if incorporation of immunotherapy
01:08:38to extrapulmonary neuroendocrine carcinoma makes
01:08:40a clinical difference.
01:08:42Terrific. And, I think one
01:08:43of the unique things about
01:08:44this study,
01:08:46often these patients come in
01:08:47with rapidly progressive disease. You
01:08:48wanna get them treated initially,
01:08:51and their LFTs may be
01:08:52elevated if they have hepatic
01:08:54metastases.
01:08:55So you can actually treat
01:08:57with one cycle of carbotoposide
01:08:59and then move on and
01:09:00randomize them on the trial.
01:09:02So, so if you've treated
01:09:04the patient, you still have
01:09:05the potential of putting them
01:09:06on study, after their initial
01:09:08induction treatment. So that's important
01:09:10to note. Yeah. Great point.
01:09:11This patient and maybe the
01:09:12next slide. He actually was
01:09:13given his first cycle of
01:09:15carboplatinetoposide
01:09:16urgently off trial
01:09:18given burden of disease.
01:09:19The trial from a prostate
01:09:21cancer perspective does allow hormonal
01:09:22therapy and second generation hormonal
01:09:24therapy orally
01:09:26to target the Gleason adenocarcinoma
01:09:28component. So he got cycle
01:09:30one off trial. He was
01:09:31started on hormonal therapy and
01:09:33enzalutamide,
01:09:34and then for cycle two
01:09:35was randomized into the atezolizumab
01:09:37arm. Got it with induction
01:09:38therapy, was randomized into the
01:09:40maintenance phase.
01:09:41So he started therapy in
01:09:43February of two thousand twenty
01:09:44four, had a very robust
01:09:45PR.
01:09:46He's eleven cycles in and
01:09:48remains on treatment today. Terrific.
01:09:50Problem is, what do we
01:09:51do with these patients afterwards?
01:09:53These patients should be molecularly
01:09:54phenotyped.
01:09:55We are moving on some
01:09:56some newer unique
01:09:58trials, in,
01:10:00this this state of disease.
01:10:02And most of them require
01:10:03that patients have at least
01:10:04a fifty percent
01:10:08component of small cell or
01:10:09neuroendocrine in their prostate cancer.
01:10:11So that's that's moving forward.
01:10:13Our final case, last but
01:10:15not least, radiation oncology. Doctor
01:10:16Ron, please,
01:10:18you have the form. Thank
01:10:19you. So this is a
01:10:20patient who had prostatectomy,
01:10:22about six, seven years ago.
01:10:24Need to salvage radiation and
01:10:25had a short term ADT
01:10:26for biochemical occurrence about three
01:10:28years ago. The PSA was
01:10:30initially undetectable after the radiation,
01:10:32but now rising to about
01:10:33one with a normal testosterone
01:10:35level still still castrate,
01:10:37sensitive.
01:10:39The updated PSMA PET scan
01:10:41showed only uptake in two
01:10:42sites, in l one and
01:10:43the sacrum.
01:10:44And and patient also personally
01:10:46prefer to try to delay,
01:10:48the ADT initiation as much
01:10:50as possible given their experience
01:10:51on the prior course.
01:10:53So in terms of next
01:10:54treatment options
01:10:59of of course, patients can
01:11:00be just beyond ADT alone.
01:11:03There's some evidence now, especially
01:11:04for maybe short
01:11:06ADT doubling time of adding
01:11:08additional antiandrogen.
01:11:10But we have this concept
01:11:11of metastases directed therapy with
01:11:13or without ADT, as an
01:11:14option. And I think it
01:11:16really speaks to the heart
01:11:17of when we have algal
01:11:18metastatic disease,
01:11:19how much microscopic
01:11:21disease there is elsewhere in
01:11:22the body that we can't
01:11:23detect.
01:11:25In this situation, we ultimately,
01:11:27in a in a multidisciplinary
01:11:29setting, elected for SBRT alone.
01:11:35And so patient received SBRT
01:11:37without ADT to the two
01:11:38sites,
01:11:39and the PSA nadir was
01:11:40undetectable a year later. And
01:11:42you can see that we
01:11:43have this typical delayed cell
01:11:45death after radiation, took about
01:11:46a year.
01:11:48Now it's about two, two
01:11:49and a half years from
01:11:50the treatment.
01:11:51The PSA is detectable,
01:11:53but,
01:11:54the patient has still not
01:11:55needed any EDT, his eugonadal,
01:11:57and has not had any
01:11:59late topsy from this, SBRT
01:12:01course.
01:12:02We're certainly very thoughtful. We
01:12:03chose a SBRT dose that
01:12:05was not the highest dose.
01:12:06This is not necessarily a
01:12:07curable situation. We wanna avoid
01:12:10compression fracture, especially with patients
01:12:12who may need long term
01:12:13insurance deprivation and risk of,
01:12:16bone density issues with that.
01:12:18No, lumbar plexopathy given the
01:12:20location of the sacral lesion.
01:12:22But, overall, you know, this
01:12:23has been a successful case.
01:12:25But it it it speaks
01:12:26to especially with such a
01:12:28great biomarker that is PSA
01:12:29without ADT,
01:12:31we we were able to
01:12:32confirm that clearly there was
01:12:33not much, if any, microscopic
01:12:35disease outside from the, two
01:12:37lesions.
01:12:38So I I really like
01:12:39this case because I think
01:12:40it highlights something that we're
01:12:41all starting to think about
01:12:42in this disease.
01:12:44The
01:12:45use of PSMA PET scanning
01:12:47earlier detecting disease earlier. This
01:12:49we don't know. Since we
01:12:50don't have conventional imaging, we
01:12:51don't know if this is
01:12:52somebody we would detect by
01:12:53a conventional scan or not.
01:12:55And then how much therapy
01:12:56is necessary
01:12:57in a patient
01:12:58who
01:12:59may be frail, may be
01:13:00elderly, may not tolerate androgen
01:13:02deprivation therapy very, very well.
01:13:04So this is why it's
01:13:05so important to have a
01:13:08discussion amongst the nuclear med
01:13:09doc who's performing the PET
01:13:11scan, the radiation oncology doc,
01:13:13and the medical oncologist,
01:13:15to be sure that we're
01:13:16doing the right thing for
01:13:17the patient. I have somebody
01:13:18like this who's ninety three
01:13:19years old,
01:13:20and, he's doing great right
01:13:22now off ADT.
01:13:24And he clearly had hot
01:13:25lesions on the PSMA PET
01:13:27scan.
01:13:28Joe, what are the,
01:13:30things that you look out
01:13:31for in a PSMA PET
01:13:32scan? Where where do you
01:13:32think you see most false
01:13:34positives?
01:13:35You know, we see a
01:13:35lot of false positive, mostly
01:13:37in the rib lesions. You
01:13:38know, we have definitely got
01:13:39the clinical context of this.
01:13:41It's kinda isolated with a
01:13:42revision.
01:13:43You know, we also look
01:13:44at the cysticordis as well
01:13:45in the absence of correlates
01:13:46where I tend to say
01:13:47this part you know, like,
01:13:49with the false positive finding.
01:13:50And we definitely get the
01:13:51lymph metastases and other pelvic,
01:13:54you know, bones as well.
01:13:55Yep. Well, I think what
01:13:57you're pointing out is that
01:13:58it's important to look at
01:13:59the overall clinical picture of
01:14:00the patient. Doctor Ahn and
01:14:01I have a patient that
01:14:02we share together
01:14:03that,
01:14:04had a relapse twelve years
01:14:06after radical prostatectomy. He lit
01:14:07up in the bed, and
01:14:08he also lit up in
01:14:09the ribs. As Joe pointed
01:14:10out Joe pointed out, there's
01:14:11a lot of false positivity.
01:14:13We gave him six months
01:14:14of hormones, radiated the bed,
01:14:16and then stopped. Those ribs
01:14:17are still there, but he's
01:14:18got a negative PSA at
01:14:19this point PSA of zero.
01:14:21So you have to put
01:14:22this in the clinical context.
01:14:24Do you think this is
01:14:25aggressive disease or not based
01:14:27upon how quickly it relapses
01:14:29and where it relapses?
01:14:30So, gentlemen, any other thoughts,
01:14:33before we wrap up for
01:14:34today?
01:14:38I'll just add, obviously, de
01:14:40novo,
01:14:40like a synchronous metastases allgable
01:14:43metastases is very different than
01:14:45the autochromous,
01:14:46allgable metastatic
01:14:47disease. We have separate trials,
01:14:49ongoing nationally, internationally in both
01:14:52settings. And the best evidence
01:14:53we have for a metastasis
01:14:55directed therapy is in the
01:14:56metagenesis algo metastatic setting right
01:14:58now.
01:14:59Exactly. In fact, there are
01:15:00trials going on right now
01:15:01with metastatic patients. We have
01:15:02one,
01:15:04SIMCAP, which is looking at
01:15:05taking doing radical prostatectomy or
01:15:07radiation therapy,
01:15:08prior to cert prior in
01:15:10a patient who has metastatic
01:15:11disease.
01:15:12It seems that they at
01:15:13least in retrospective data, the
01:15:15treatment of primary does have
01:15:17a different impact on overall
01:15:18survival.
01:15:19Mike, you Yeah. No. I
01:15:20just think that because PSMA
01:15:22PET is so sensitive,
01:15:24we feel a little more
01:15:24comfortable if we see a
01:15:25lesion and we don't see
01:15:26a lot of other stuff
01:15:27to ask doctor Ahn and
01:15:29his colleagues to go after
01:15:30things that we never would
01:15:31have done in the past.
01:15:33And probably,
01:15:34for a lot of patients,
01:15:35can sort of keep them
01:15:36going for a long time
01:15:37without having to worry so
01:15:38much about other things even
01:15:39if it may not be
01:15:39curable. It gives them durable
01:15:41responses for a long time.
01:15:42And I would just extrapolate
01:15:44on case three. The other
01:15:44cyst
01:15:46situation where I've had really
01:15:47good success with this is
01:15:48elderly frail people
01:15:50that are becoming progressively castrate
01:15:52resistant that I don't wanna
01:15:53escalate care on or treatment
01:15:54on. Mhmm. So I have,
01:15:55an eighty seven year old
01:15:57who's now progressing on Zytiga,
01:15:58had oligometastatic progression,
01:16:00really not a great chemotherapy
01:16:02candidate, and we SBRT'd. He
01:16:03had single site progression
01:16:05and has done fantastic even
01:16:07at time of progression with
01:16:09the goal of delaying escalation
01:16:10of systemic therapy.
01:16:12Right.
01:16:13Gentlemen, thank you for your
01:16:15cases and comments. I'm going
01:16:16to point doctor doctor.
01:16:29Great. Thank you to the
01:16:30GU team. We are next
01:16:32moving to GI.
01:16:34I'm not allowed to have
01:16:35favorites, but this might be
01:16:37might be my favorite. Doctor
01:16:38Lacey, you can come over.
01:16:40So, so we will kick
01:16:41this off with
01:16:43a hot topics.
01:16:44I'll move for you. Hot
01:16:45topics in, a GI cancers
01:16:48by doctor Jill Lacy,
01:16:49and then we will move
01:16:50to a case discussion that
01:16:52I will moderate,
01:16:53with presenters doctors Laura Baum,
01:16:55Jeremy Kormansky,
01:16:57Saj Khan, and Kevin Du.
01:16:59Thanks, Jill.
01:17:08Okay. So thank you, Pam,
01:17:10and welcome everyone this morning.
01:17:11It's truly a pleasure to
01:17:13be here today speaking on
01:17:14the hot topics in GI
01:17:16cancer, and here are my
01:17:17disclosures, and here is the
01:17:18agenda.
01:17:19So I have selected a
01:17:20few of many, I think,
01:17:22impactful
01:17:23oral presentations at ASCO and
01:17:25ESMO from twenty twenty four,
01:17:27starting with resectable esophageal adenocarcinoma,
01:17:29the ISApec trial, and then
01:17:31moving to metastatic colorectal cancer,
01:17:33advanced HCC, advanced to render
01:17:35consumers.
01:17:36And if time permits, it
01:17:37probably won't, I'd like to
01:17:38touch on some of the
01:17:39promising new therapeutics that are
01:17:41really bringing
01:17:42hope to this field and
01:17:43where Yale has had, an
01:17:45important role,
01:17:46in studying these drugs.
01:17:49So here are the key
01:17:50takeaways from each of these
01:17:52presentations,
01:17:52in bulleted form, and I'll
01:17:54highlight these as we go
01:17:55along.
01:17:55I think the overarching,
01:17:57takeaway from my perspective is
01:17:59that in twenty twenty four,
01:18:00we have made significant progress
01:18:02in GI cancers,
01:18:03albeit, admittedly, it is more
01:18:05incremental,
01:18:06than transformational.
01:18:08But we have ever more
01:18:09options for our patients,
01:18:12and so careful patient selection
01:18:14and weighing the toxicities against
01:18:16the efficacy is increasingly
01:18:18important
01:18:19in outlining a long term
01:18:21management plan for our patients.
01:18:24So we'll start with Isopec,
01:18:26perioperative chemotherapy FLOT versus neoadjuvant
01:18:29chemoradio radiotherapy or CROS
01:18:31for patients with resectable esophageal
01:18:33adenocarcinoma.
01:18:35So Isopec was a plenary
01:18:36presentation. It's received much attention,
01:18:39and it has been herald
01:18:40as practice changing. So we're
01:18:42gonna take a careful look
01:18:43at what we learned from
01:18:44Isopec and also importantly what
01:18:46we didn't learn.
01:18:47So why do Isopec? So
01:18:49in twenty sixteen, when this
01:18:51study was initiated,
01:18:52we had two options for
01:18:53this patient population.
01:18:55Both had been shown to
01:18:56be superior to esophagectomy
01:18:58alone in these three randomized
01:19:00trials
01:19:01highlighted here in this figure.
01:19:03So we, most often, we're
01:19:05using preoperative chemoradiotherapy
01:19:07with the cross regimen widely
01:19:08used, which was radiation with
01:19:10low dose weekly carboplatin and
01:19:12paclitaxel.
01:19:14We also had the option
01:19:15of a nonradiotherapy
01:19:16treatment with perioperative chemotherapy
01:19:19with the FLOP regimen, flouriericiloxaliplatin,
01:19:21and,
01:19:22Taxotere
01:19:23shown to be superior to
01:19:25the very old,
01:19:26MAGIC,
01:19:27trial.
01:19:28So there was no clear
01:19:29winner between these two options,
01:19:30and part of that was
01:19:31because these randomized phase three
01:19:33trials included a very mixed
01:19:34patient population.
01:19:36The perioperative trials included gastric
01:19:38adenocarcinoma,
01:19:39actually were predominantly gastric cancer
01:19:41studies, and the CROS study
01:19:43included squamous cell histology, which
01:19:45is really a very different
01:19:46disease biologically
01:19:47we know.
01:19:48So I think when we
01:19:49put ESAPEC in context in
01:19:51twenty twenty four, it's important
01:19:53to appreciate the treatment evolution
01:19:54that we've seen since this
01:19:55study started and largely really
01:19:57since twenty twenty one, and
01:19:59I've highlighted the three, publications
01:20:01that are relevant to this.
01:20:02So we've shifted away in
01:20:04the adenocarcinoma
01:20:05patient population
01:20:06from CROS
01:20:08to Pofax, and that's based
01:20:10on the disappointing long term
01:20:11results of CROS in adenocarcinoma.
01:20:14It's better in squamous cell
01:20:15carcinomas
01:20:16and the very encouraging data
01:20:18with FOLFOX using more induction
01:20:19FOLFOX as well as radiation
01:20:21from the CLGB eight zero
01:20:22eight zero three study.
01:20:24We now, routinely use adjuvant
01:20:26nivolumab. This is a standard
01:20:28of care since twenty twenty
01:20:30one after preoperative
01:20:31chemoradiotrathy
01:20:32in the high risk patient
01:20:33population with residual carcinoma and
01:20:35esophageectomy spec specimen.
01:20:38Increasingly, we are considering organ
01:20:40preservation in selected patients with
01:20:42the complete clinical response. This
01:20:44is what's happened in rectal
01:20:45cancer,
01:20:46again, where the morbidity of
01:20:47surgery is high. And then
01:20:48finally, for the MSI high
01:20:50patient population, which is about
01:20:52five percent of our patients,
01:20:54we go down a completely
01:20:55different treatment paradigm incorporating immune
01:20:57checkpoint inhibitors in the neoadjuvant
01:20:59setting.
01:21:01So here's the ISApec trial
01:21:02scheme. It's very straightforward,
01:21:04a one to one randomization
01:21:05of, perioperative flat versus cross,
01:21:09with a locally advanced patient
01:21:11population,
01:21:12muscle invasive or node positive
01:21:14adenocarcinoma,
01:21:15no metastases.
01:21:16Importantly,
01:21:17patients had to be medically
01:21:19fit and, I would say,
01:21:19also agree to an esophagectomy.
01:21:22Gastric cancer patients and squamous
01:21:24cell patients were excluded. So
01:21:25this is a more homogeneous
01:21:27clean population,
01:21:28primary endpoint overall survival.
01:21:31It's important to appreciate that
01:21:32in the cross arm, a
01:21:33third of the patients did
01:21:34actually not complete chemoradiation therapy.
01:21:36That's a surprise in and
01:21:38a much higher percentage than
01:21:39we've seen in other studies.
01:21:40Not surprising was that about
01:21:42forty seven percent of patients
01:21:44did not complete postoperative FLOT.
01:21:46It's very hard to give
01:21:47chemotherapy
01:21:48after an esophagectomy.
01:21:50So here are the results.
01:21:51This was a positive study
01:21:52favoring FLOT, with an improvement
01:21:54in,
01:21:55overall,
01:21:56survival with a p value
01:21:58or hazard ratio of point
01:21:59seven.
01:22:00There was also an improvement
01:22:01in progression tree survival.
01:22:03And the additional findings were
01:22:05that both of these approaches
01:22:06led to similar R0 resection
01:22:08rates. They were both reasonably
01:22:10safe.
01:22:11What was a little disappointing
01:22:13was the low pathologic CR
01:22:14with both treatments, so these
01:22:15are not really great strategies
01:22:17for organ preservation.
01:22:19And also disappointing is that
01:22:20still we are not curing
01:22:22about fifty percent or more
01:22:23of patients. And so clearly,
01:22:25we need better systemic therapies
01:22:27because most of the relapse
01:22:28relapses are metastatic.
01:22:31So importantly, we should not
01:22:33extrapolate ESApEC to patients who
01:22:35are not proceeding to an
01:22:36esophageectomy.
01:22:37Those patients should get chemo
01:22:38radiotherapy,
01:22:39not to squamous cell cancers,
01:22:41and not to MSI high
01:22:42tumors.
01:22:43So we know what we
01:22:44learn, but what did we
01:22:46not learn? And I think
01:22:47this is important. So the
01:22:48elephant in the room right
01:22:49now is what about immunotherapy?
01:22:51So the standard of care
01:22:53has really moved to preoperative
01:22:54chemoradiotherapy
01:22:56followed by eduvant nivolumab if
01:22:57there's not a path CR.
01:22:59That's based on CheckMate five
01:23:00seventy seven where we have
01:23:02impressive PFS benefit. No overall
01:23:04survival data yet.
01:23:06So should we just add
01:23:07immunotherapy to FLAT extrapolating from
01:23:09CheckMate five seventy seven? I
01:23:10think the answer to that
01:23:11is clearly no. There are
01:23:13three studies looking at this,
01:23:14and the keynote five eighty
01:23:16five study adding pembro to
01:23:17perioperative chemotherapy was a negative
01:23:19study, so we should not
01:23:20be doing this yet. And
01:23:22then what is the role
01:23:23of preoperative chemotherapy when it's
01:23:24added to optimum systemic therapy?
01:23:26Which cross is not?
01:23:28And I think this question
01:23:29was,
01:23:30partially,
01:23:32but not really addressed in
01:23:33the Top Gear study.
01:23:36So we have waited many,
01:23:37many long years to get
01:23:38the results of Top Gear,
01:23:40which was asking the question,
01:23:41is there a benefit to
01:23:42neoadjuvant chemoradiotherapy
01:23:44added to standard perioperative chemotherapy?
01:23:47Now this is really a
01:23:48gastric cancer trial. You can
01:23:49see that in the title.
01:23:51This was presented at ASCO
01:23:52and published simultaneous time simultaneously
01:23:55the same day in the
01:23:55New England Journal of Medicine.
01:23:57So take that under advisement.
01:23:58Esophageal adenocarcinomas
01:24:00were excluded. There were a
01:24:01subset of patients with EGJ
01:24:03cancers.
01:24:04So cut to the chase.
01:24:05The bottom line is that
01:24:07adding chemorated therapy to perioperative
01:24:09chemotherapy and gastric cancer,
01:24:11is not a benefit, although
01:24:13it does not increase toxicity,
01:24:14and it does increase the
01:24:16pathologic complete response and leads
01:24:18to more tumor downstage but
01:24:19not translating to survival benefit.
01:24:22So perioperative
01:24:24chemotherapy remains the standard of
01:24:25care in gastric cancer. No
01:24:27chemo radiotherapy. No immune checkpoint
01:24:28inhibitors. I think the jury
01:24:30is out for esophageal adenocarcinoma
01:24:32as to whether there's a
01:24:33subset of patients that might
01:24:34benefit from chemo radiotherapy. And
01:24:36in fact, Top Gear actually
01:24:38suggested that in the subset
01:24:39analysis
01:24:40with a hazard ratio of
01:24:41point seven eight,
01:24:43for chemoradiotherapy
01:24:44in the EGJ patients.
01:24:46So here we are twenty
01:24:47twenty four, and this slide
01:24:48summarizes what I just went
01:24:49through. The pendulum swings,
01:24:52cross out cross,
01:24:53FLOTS in. I think the
01:24:54story certainly will be continued,
01:24:57but FLOT is the preferred
01:24:58treatment for now. Nivolumab in
01:25:00the adjuvant setting should be
01:25:02limited to patients who were
01:25:03treated with, chemoradiotramp
01:25:06neoadjuvant leak.
01:25:07I think we still need
01:25:08to understand how to deploy
01:25:10chemoradiotrampy
01:25:11with good systemic therapy and
01:25:12benefits.
01:25:13And then importantly, for patients
01:25:15who are declining, deferring, or
01:25:16cannot undergo surgery,
01:25:18chemoradiotrampy
01:25:19is still a part of
01:25:19the standard of care, preferably
01:25:21with more and better systemic
01:25:23therapy.
01:25:24So we're gonna move now
01:25:25to metastatic colorectal cancer, and
01:25:27there were two really important
01:25:29presentations at ASCO
01:25:31looking at patients with liver
01:25:33only,
01:25:34stator colorectal cancer. So the
01:25:36first was the collision trial.
01:25:38This trial was focused on
01:25:39patients with receptible liver only
01:25:41metastatic disease. So I think
01:25:43we've all known for about
01:25:44four decades now that hepatic
01:25:46resection
01:25:47for patients with liver only
01:25:48metastatic CRC,
01:25:50is curative. Although with long
01:25:51term survival, the cure rate
01:25:52drops to only about twenty
01:25:53or twenty five percent of
01:25:54patients, but curable nonetheless.
01:25:56And the big question in
01:25:57the last decade or more
01:25:59has been whether,
01:26:01surgical resection is better than
01:26:03taking a less invasive ablation
01:26:05approach with microwave or radiofrequency
01:26:07ablation.
01:26:08And the retrospective studies have
01:26:09yielded mixed conclusions.
01:26:11In NCCN, hepatic resection still
01:26:13is sort of the gold
01:26:14standard treatment of choice with
01:26:15thermal ablation
01:26:16consideration.
01:26:18So the collision trial was
01:26:20designed to test the hypothesis
01:26:22that thermal ablation is noninferior
01:26:24to surgery. So here is
01:26:26the design.
01:26:27Patients with limited ten or
01:26:29less metastases,
01:26:31were randomized
01:26:33to resection.
01:26:34Ablation was also permitted if
01:26:36needed
01:26:37or ablation with limited resection
01:26:39also permitted.
01:26:40Most of these patients had
01:26:41a limited tumor burden, three
01:26:43or fewer small lesions.
01:26:45And, again, this was a
01:26:46non inferiority study.
01:26:49So this study met its
01:26:50primary endpoint. Thermal ablation was
01:26:52non inferior
01:26:53to surgery in this study.
01:26:55No difference in survival.
01:26:57Also, no difference in local
01:26:58and distant progression free survival,
01:27:00but differences in adverse events,
01:27:02length of hospital stay favoring
01:27:04thermal ablation.
01:27:06So thermal ablation, I think,
01:27:07clearly now is a comfortable
01:27:09option in selected patients with
01:27:11small limited liver only metastatic
01:27:12colorectal cancer, I think we
01:27:14can feel comfortable that we
01:27:15are probably not compromising cure.
01:27:17Although it's important
01:27:18to follow this these these
01:27:20data long term,
01:27:22given the late relapses that
01:27:23we can see. I think
01:27:25the question is how generalizable
01:27:26is this strategy given the
01:27:28patient selection that went into
01:27:29the study.
01:27:30I'm sure there are gonna
01:27:31be strong institutional biases and
01:27:33certainly differences in institutional expertise.
01:27:35And I think for the
01:27:35oncologists in the room, the
01:27:37big question is always how
01:27:39much and what systemic therapy
01:27:40is needed both before and
01:27:42after, especially if you have
01:27:43been in the high rate
01:27:44of relapse.
01:27:45So now we're gonna pivot
01:27:47to patients with unresectable liver
01:27:48only metastases. And, unfortunately,
01:27:51most patients with liver only
01:27:52disease will not be candidates
01:27:54for consideration of a collision
01:27:55strategy.
01:27:56They have multiple disease in
01:27:57both lobes. And we've had
01:27:59intriguing data regarding transplant,
01:28:01liver transplant in this,
01:28:03population. So the TransMet study
01:28:06was designed to address the
01:28:07role of transplant in these
01:28:08patients.
01:28:10So here is the trial
01:28:11design. Patients were randomized to
01:28:13transplant plus chemo versus chemo.
01:28:16Very selected patient population,
01:28:18younger, good performance status, no
01:28:20BRAF, lower c e CEA.
01:28:23And these patients went through
01:28:24really a very rigorous selection
01:28:26and prioritization
01:28:27process.
01:28:28So about half the patients,
01:28:30were deemed ineligible
01:28:32after independent multidisciplinary
01:28:34review. So ninety four patients
01:28:36were randomized and seventy four
01:28:38completed
01:28:38therapy per protocol.
01:28:40So these results were impressive.
01:28:42Obviously,
01:28:43the, Kaplan Meier curve speak
01:28:45for themselves with a significant
01:28:46improvement in five year overall
01:28:48survival, the primary endpoint in
01:28:50both intent to treat and
01:28:51per protocol patients.
01:28:53This is not a cure.
01:28:55Seventy two percent of patients
01:28:56who were transplanted did recur,
01:28:58but many of those recurrences
01:28:59were oligometastatic
01:29:00lung and could be handled
01:29:02with focal therapies.
01:29:03So the takeaway here,
01:29:05in the context of very
01:29:07rigorous patient selection,
01:29:09Liver transplant, I think, offers
01:29:11the hope
01:29:12of long term survival in
01:29:13this very difficult patient population,
01:29:16compared to chemotherapy alone.
01:29:19So this is very tantalizing
01:29:20and very exciting, but it
01:29:22has it that has really
01:29:23led to to many questions
01:29:24and many challenges, and I've
01:29:25highlighted a few of them
01:29:26here.
01:29:27And this list could be
01:29:28much longer, but I think
01:29:29I would like to jump
01:29:30to the bottom. And that
01:29:31is, can we, ever achieve
01:29:33equitable equitable access, at least
01:29:35in the United States, to
01:29:36transplant for appropriate patients. And,
01:29:39I stole this figure from
01:29:40the discussant,
01:29:42of this study who showed
01:29:44the distances that patients in
01:29:45the United States are traveling
01:29:47to access liver transplantation.
01:29:49And so I think, it
01:29:50goes without saying that many
01:29:51of our patients will be
01:29:52excluded,
01:29:53from this very promising treatment.
01:29:56So we're gonna flip now
01:29:58to,
01:30:00microsatellite instability high, metastatic
01:30:03colorectal cancer. So going into
01:30:05twenty twenty four,
01:30:06the standard of care here
01:30:08was pembro
01:30:09based on the keynote one
01:30:10seventy seven study, which showed
01:30:12dramatically superior PFS and OS
01:30:14with pembro
01:30:16versus chemo in the first
01:30:17line setting for MSI high
01:30:18metastatic colorectal cancer.
01:30:20So at ASCO GI and
01:30:22ASCO, we heard the results
01:30:23from CheckMate eight h w,
01:30:27which had a three arms
01:30:28in it,
01:30:29standard of well, FOLFOX, not
01:30:31standard of care,
01:30:33nivo plus ipi and nivo.
01:30:36And we heard about the
01:30:37results from the chemo versus
01:30:39ipinivo
01:30:40arms. We are eagerly awaiting
01:30:42the most important results from
01:30:44the nivo versus ipinivo arms.
01:32:17This is in the advanced
01:32:18HCC space. We're moving from
01:32:20CheckMate eight h w to
01:32:22nine d w. Good luck
01:32:24keeping those straight.
01:32:25So this was a randomized
01:32:27trial in advanced,
01:32:30good patients with HCC, so
01:32:32child's Pua,
01:32:34good ecard performance
01:32:35status,
01:32:36comparing a TKI, lenabasirafenib,
01:32:39against the nivo ipi combination
01:32:41with the higher dose of
01:32:42ipi,
01:32:43primary input overall survival.
01:32:45And this was a positive
01:32:46study for overall survival, hazard
01:32:48ratio point seven nine. I
01:32:50think it's important to look
01:32:51at the treatment related adverse
01:32:53events, twenty five percent serious
01:32:55and four percent deaths. That's
01:32:56a little bit sobering.
01:32:58And, of course, again, prostrale
01:33:00comparison is inevitable
01:33:02with the two other regimens
01:33:03we have,
01:33:05atezo bev and the Stride
01:33:07regimen with a single dose
01:33:08of tremie and dervo.
01:33:10And, again, sort of you
01:33:11be the judge here, but
01:33:12what's notable about ipinivo is
01:33:14the response rate and the
01:33:15durability of response. So, again,
01:33:17another first line option in
01:33:19this setting,
01:33:20in addition to atezoBab and
01:33:22the STRIDE regimen.
01:33:23And, again, for those patients
01:33:24who are not eligible for
01:33:26BAB, which is a considerable
01:33:27number due to varices
01:33:29and where response may be
01:33:30a priority, and, again, balancing
01:33:32toxicity against efficacy.
01:33:34And last but not least,
01:33:36a very important study in
01:33:38advanced neuroendocrine tumors.
01:33:40This is the cabinet study
01:33:42that was presented in an
01:33:43oral presentation
01:33:44at ESMA with a simultaneous
01:33:46publication in New England Journal
01:33:48of Medicine, very much deserved
01:33:49in my opinion.
01:33:50So the patient population here
01:33:52is,
01:33:53well differentiated neuroendocrine tumors from
01:33:56any site, including lung.
01:33:58Patients had to have had
01:33:59a prior, not only somatostatin
01:34:01analog, but have progressed on
01:34:03one other FDA approved treatment.
01:34:05So these are patients with
01:34:06advanced neuroendocrine tumors refractory
01:34:09to at least one line
01:34:10of treatment.
01:34:11Sixty percent of the patients
01:34:13notably had had prior Lutathera.
01:34:14So I think this is
01:34:15important, really important data for
01:34:17that patient population,
01:34:19and the majority of the
01:34:20patients had prior everolimus.
01:34:22So here's the trial design,
01:34:23extra pancreatic and pancreatic,
01:34:25two to one Cabo versus
01:34:26placebo,
01:34:27primary endpoint progression free survival
01:34:29by blinded central review. I
01:34:31think these results are really
01:34:33impressive. This is very exciting
01:34:34from my opinion,
01:34:36with CABO significantly improving the
01:34:38PFS in previously treated patients,
01:34:40not only in pancreatic neuroendocrine
01:34:42tumors, which I think we
01:34:43might expect, but also in
01:34:45the extra pancreatic tumors, including
01:34:46lung. And benefit was seen
01:34:48in all subgroups.
01:34:49And so this is really
01:34:50the first reported randomized study
01:34:52looking at treatment options after
01:34:54prior Lutathera or other targeted
01:34:56drugs. And, again,
01:34:58another treatment option for patients
01:35:01with previously treated neuroendocrine tumors.
01:35:03So the challenge that doctor
01:35:04Coontz faces every day in
01:35:05clinic is with these expanding
01:35:07choices, including
01:35:08liver directed therapies,
01:35:10what is the optimum sequencing
01:35:12of these modalities? And I
01:35:13think that is a story
01:35:14to be continued.
01:35:16In the interest of time,
01:35:17I will not go through
01:35:19these promising new therapeutics. I
01:35:21will show them here, though.
01:35:22Dornvanalumab,
01:35:24which is a TIGID inhibitor,
01:35:25which is showing very encouraging
01:35:26promise in treatment naive metastatic
01:35:29esophageal
01:35:30cancers in combination with the
01:35:31PD one inhibitor.
01:35:33We saw that data. An
01:35:35adenosine receptor inhibitor from ARCUS,
01:35:38looking very promising in combination
01:35:40with bevacizumab and FOLFOX and
01:35:41chemo refractory colon cancer.
01:35:44And I think very, very
01:35:45encouraging data
01:35:48with AbbVie's
01:35:49a d four hundred antibody,
01:35:52to a conjugate to CMAT,
01:35:55which is showing very impressive
01:35:56activity in a highly refractory
01:35:58colorectal cancer patient population. So
01:36:01to be continued, and these
01:36:02have been,
01:36:03drugs that we have had
01:36:04in trials here at Yale
01:36:05and continue to do so.
01:36:07So we back to our
01:36:08key takeaways. We'll conclude with
01:36:10that. I think this is
01:36:11a very exciting time, not
01:36:13only for increasing options
01:36:15for our patients, but also
01:36:17I think the promise of
01:36:18new, therapeutics. I think we
01:36:20are truly entering a new
01:36:21era,
01:36:22in the GI oncology space.
01:36:25So thank you all for
01:36:26listening through this, and I
01:36:28I,
01:36:31I,
01:36:32I have to acknowledge ASCO
01:36:34and ESMO for making all
01:36:35of this really exciting data
01:36:37publicly available for me to
01:36:38borrow and, most importantly, to
01:36:40the presenters, the authors, and
01:36:41the discussants,
01:36:43for their slides and some
01:36:44of their content,
01:36:45and to my GI oncologist
01:36:47who gave me a lot
01:36:48of feedback about how to
01:36:50contextualize
01:36:50these, studies.
01:37:00Great. I would love to
01:37:01invite the panelists for our
01:37:04case discussion up, so doctors
01:37:06Baum, Du, Khan,
01:37:08Kormansky.
01:37:09Doctor Lacey, you're welcome to
01:37:10stay up if you would
01:37:11like.
01:37:28Great. So we are we
01:37:29have two cases, and we
01:37:30are going to try to,
01:37:32emphasize a few of the
01:37:33points that doctor Lacey made
01:37:35a little bit earlier with
01:37:36a few of the,
01:37:37abstracts.
01:37:39Okay. The first is on
01:37:40esophageal cancer.
01:37:42So this is a seventy
01:37:43four year old who presented
01:37:44with a three month history
01:37:45of dysphagia to solids associated
01:37:48with regurgitation
01:37:49and a thirty pound intentional
01:37:52says he said, weight loss
01:37:53over eight months. He had
01:37:55a history of melanoma of
01:37:56the shoulder, a status post
01:37:58resection that was t three
01:37:59b sent in a lymph
01:38:00node negative four years prior,
01:38:03other medical conditions as listed.
01:38:06And,
01:38:07with, medications listed here, he
01:38:09was married, has was just
01:38:11getting ready to move out
01:38:11of the state. His wife
01:38:13also had a recent history
01:38:14of a cancer diagnosis,
01:38:16no alcohol or tobacco,
01:38:18had a golden retriever and
01:38:20no
01:38:21children. So, what is the
01:38:23recommended workup? And I'm gonna
01:38:25let doctor Baum help us,
01:38:26think about that.
01:38:28Thanks, Pam. So, obviously, for
01:38:30esophageal cancer, we'll need a
01:38:32biopsy.
01:38:33To complete staging, we'll need
01:38:35the e EEGD with EUS.
01:38:37I think what's become complicated
01:38:39or maybe less complicated, but
01:38:40intellectually complicated to think through
01:38:43is the nodal staging.
01:38:45So are you looking for
01:38:46regional versus versus distant nodes?
01:38:49Until two thousand ten,
01:38:51you know, a lot of
01:38:51it had to do with
01:38:52where the primary tumor was
01:38:53versus the regional tumor.
01:38:55But since then, we've been
01:38:57staging regardless of the regional
01:38:59site as long as there's
01:39:01one to two nodes that's
01:39:02n one, three to six
01:39:03is n two,
01:39:04and seven plus that's n
01:39:05three. So you're really thinking
01:39:07of nodal staging
01:39:08from, TNM by that valuation,
01:39:11but complicating that with how
01:39:13you're considering the radiation field
01:39:14and the surgical field and
01:39:16then also supraclavicular
01:39:18nodes, for example, would be
01:39:20would still be distant.
01:39:21So in order to try
01:39:23to,
01:39:24make sure you avoid,
01:39:26an unnecessary
01:39:27major surgery,
01:39:29we are getting PET CTs
01:39:31in all of our esophageal
01:39:32cancer patients, which is upstaging
01:39:34about
01:39:35somewhere between
01:39:37ten five to ten to
01:39:38twenty percent of patients.
01:39:40And,
01:39:41I at least am referring
01:39:43many or all of patients
01:39:45with t three or t
01:39:46four,
01:39:47esophageal
01:39:48gastric junction disease
01:39:50to doctor Khan's team, for
01:39:52staging,
01:39:54laps.
01:39:56So,
01:39:57there's a few things up
01:39:58there as well, and I
01:39:59just wanna highlight the multidisciplinary
01:40:00tumor board evaluation,
01:40:03keeping in mind that it
01:40:05is a complicated decision now
01:40:07balancing
01:40:08the
01:40:09whether you're starting with chemotherapy,
01:40:11chemo radiotherapy,
01:40:13and whether the patient is
01:40:14considering,
01:40:15organ preservation
01:40:16sort of nonsurgical option
01:40:18or is definitively going to
01:40:20surgery.
01:40:22Great. Thank you, Laura.
01:40:25So this patient went on
01:40:26to have,
01:40:27this diagnostic workup. So his
01:40:29labs were normal. He had
01:40:31an EGD that showed a
01:40:32mass noted at the lower
01:40:33third of the esophagus between
01:40:35thirty seven to forty centimeters
01:40:37that was partially obstructing.
01:40:39The biopsy showed a moderately
01:40:41differentiated adenocarcinoma
01:40:43that was HER2 three plus.
01:40:45Note, this was a couple
01:40:46of years ago, and so
01:40:47we did not have some
01:40:48of the standard,
01:40:51immunohistochemical
01:40:52or or, genetic markers that
01:40:54we normally do. This is
01:40:55a very busy slide, but
01:40:57I am going to turn
01:40:57to doctor Kourtmansky just to
01:40:59have us comment a little
01:41:00bit on how
01:41:01our molecular testing for GI
01:41:03pathology has evolved. And this
01:41:04is, I will say, major
01:41:06credit
01:41:07to doctors Gibson and Zuchen
01:41:09Zhang as a partnership with
01:41:10PATH.
01:41:12So I think,
01:41:14one of the things that's
01:41:15been clear over the last
01:41:17several years is that,
01:41:21although cytotoxic chemotherapy still remains
01:41:23the backbone of of what
01:41:24we do,
01:41:26that there are a lot
01:41:28of molecular targets that are,
01:41:30that have been demonstrated to
01:41:32be beneficial
01:41:33across the disease spectrum. And
01:41:35so, for patients
01:41:37with upper GI malignancies,
01:41:40we think about,
01:41:41microsatellite instability. In fact, we
01:41:43think about microsatellite
01:41:45instability
01:41:46in all patients with GI
01:41:47malignancies, and this should be,
01:41:50a test that is done
01:41:51de facto for a new
01:41:52diagnosis.
01:41:54But in upper GI malignancies,
01:41:55we also think about HER2
01:41:57expression.
01:41:59We think about,
01:42:01PD L1 expression.
01:42:03And recently, we've been thinking
01:42:05about claudine expression as well,
01:42:07which is a new target,
01:42:09seen in,
01:42:11GE junction and gastric cancers.
01:42:13There's not yet an approved
01:42:15therapy for that particular target,
01:42:17although there is,
01:42:18promising data,
01:42:20with a new antibody treatment
01:42:22in combination with chemotherapy, and
01:42:23so we expect that approval
01:42:26at some point. We've been
01:42:27saying that for, like, six
01:42:28months Yeah. Now. But it's
01:42:30coming. It's coming.
01:42:32Great.
01:42:33But then, you know, HER2
01:42:35is also relevant in biliary
01:42:36tract cancers. It's relevant in
01:42:38colorectal cancers.
01:42:40And so we you know,
01:42:42as this slide outlines, we
01:42:43have a lot of testing
01:42:44that we do de facto,
01:42:45but we really do encourage,
01:42:48molecular
01:42:49profiling,
01:42:50for all patients with new
01:42:52blood GI malignancies.
01:42:54Great. Thanks, Jeremy.
01:42:57So moving on with this
01:42:59case. So an FDG PET
01:43:00CT showed an avid irregular
01:43:03circumferential thickening involving lower esophagus
01:43:06and likely GE junction. There
01:43:07was no metastatic disease.
01:43:09On endoscopic ultrasound, the patient
01:43:11was staged as a clinical
01:43:13t two.
01:43:14Though per doctor Boffa, if
01:43:16the patient has dysphagia, it's
01:43:17t three.
01:43:19And so how do we
01:43:20think about treatment?
01:43:22So I'd like to,
01:43:23I'm gonna turn to doctors
01:43:25Du and doctor Lacey. And
01:43:26and I'll the context is
01:43:27that we actually had a
01:43:28robust debate in one of
01:43:30our seminars
01:43:31on the ISOPEC trial. I'm
01:43:32gonna let doctor Du start
01:43:33since Jill had a little
01:43:34bit of time already.
01:43:36How do you think about
01:43:37this in terms
01:43:38of radiation,
01:43:39chemotherapy, and what our options
01:43:41are? And I'm gonna put
01:43:41back up this table that
01:43:43helps us think about this.
01:43:44Sure. So I'll say I'm
01:43:46a I'm
01:43:47Yeah. K. I I'm I'll
01:43:49say that I'm, I'm kind
01:43:51of a a follow-up back
01:43:52to doctor Lacey's incredible,
01:43:54discussion already about ESOPEC. And,
01:43:56you know, for this is
01:43:57a GE junction
01:43:59cancer.
01:44:00The issue with GE junction
01:44:01cancers is that they've been
01:44:03folded into many esophageal cancer
01:44:06trials, as well as gastric
01:44:08cancer trials.
01:44:09And clearly, as the the
01:44:10TOPGURE trial,
01:44:12disappointingly for me demonstrates, is
01:44:14that, there really is a
01:44:16limited role of radiation for
01:44:17gastric cancer in the modern
01:44:19era,
01:44:20for resectable cancers.
01:44:22But for esophageal cancers, I
01:44:23think the role of radiation
01:44:25still remains very relevant, and
01:44:27the
01:44:28benefits of radiation
01:44:30is still the same as
01:44:31it always was, which is
01:44:33that of improved local control,
01:44:35improved downstaging,
01:44:37and, the, very promising possibility
01:44:40and ever increasing rates of
01:44:42pathologics,
01:44:43complete response rates with, with
01:44:45chemoradiation.
01:44:47And, it is still reasonable
01:44:49in this case despite the,
01:44:51Esopec trial given the concerns
01:44:53with the,
01:44:55unbalanced chemotherapy
01:44:56given in the arms,
01:44:58to consider,
01:45:00the chemoradiation.
01:45:01As you,
01:45:02said, Pam, this patient was
01:45:04treated a few years ago.
01:45:05And, you know, I think
01:45:07at that time, even when
01:45:08you were treating this, it
01:45:10there was an awareness that,
01:45:12the low dose carboxyl
01:45:14chemotherapy given with the CROS
01:45:16arm,
01:45:17you know, the chemo radiation
01:45:19arm of the ISOBEC trial,
01:45:21was not the right chemotherapy
01:45:23to give with, radiation, and,
01:45:25we can talk about that.
01:45:27Anything else to add, Jill?
01:45:29Can you comment maybe briefly
01:45:31on the c h CALGB
01:45:33eight zero eight zero three?
01:45:35Jill, if you wanna do
01:45:36that. Yeah. So
01:45:41Yeah. So I think I
01:45:42I share, the bias and
01:45:45view that that,
01:45:46we just heard, from Kevin.
01:45:49You know, I think, we
01:45:50we know that that CROS
01:45:52does not offer effective systemic
01:45:54therapy, and patients are largely
01:45:56recurring with metastatic disease.
01:45:58I think what CLGB
01:45:59showed us is that when
01:46:01we include
01:46:02more effective systemic therapy so
01:46:05that was FOLFOX,
01:46:06three cycles of induction FOLFOX,
01:46:08and then three cycles of
01:46:09FOLFOX with chemotherapy.
01:46:11So six cycles, which is
01:46:12actually more than what patients
01:46:14are often getting with perioperative
01:46:16chemotherapy because we can't give
01:46:18the chemo after surgery.
01:46:21Those results were very impressive
01:46:23with overall survival rate, rates
01:46:25that that rival anything that
01:46:26we've seen.
01:46:28And the advantage of that
01:46:29in terms of starting with
01:46:31induction FOLFOX is that often
01:46:33pretty rapidly alleviates dysphagia
01:46:35It makes tolerability of the
01:46:36chemo radiation therapy better. I
01:46:38think impressive pathologic complete response
01:46:41rates, which is extremely important
01:46:43given that many, many patients
01:46:45are really seeking organ preservation
01:46:47now, learning about the long
01:46:48term morbidity of an esophagectomy.
01:46:50And so this is a
01:46:51shift that we have gone
01:46:52through in rectal cancer, and
01:46:54I think now are seeing
01:46:55in esophagus.
01:46:56Yep. Great. Thank you.
01:46:58So this patient went on
01:46:59to receive multimodality
01:47:01treatment as per the CLGB
01:47:02trial. Again, this was pre
01:47:04ESOPEC data. So received a
01:47:06FOLFOX and then concurrent FOLFOX
01:47:08with radiation therapy.
01:47:10I'll just put in a
01:47:11plug that DPT
01:47:12DPD testing was done prior
01:47:14to the start of fibrofib,
01:47:15which was negative. We can
01:47:16certainly address questions on that,
01:47:17but that has become part
01:47:18of standard of care for
01:47:20fluoro pyrimidine based treatments.
01:47:22Kevin, for sake of time,
01:47:23I may not go through
01:47:24you you you can do
01:47:24you wanna you a couple
01:47:26comments briefly on your pretty
01:47:27pictures?
01:47:30Since you invite me, that's
01:47:32I'll I'll just comment very
01:47:33briefly. This is, an example
01:47:34of our radiation plans,
01:47:36showing really very conformal distribution,
01:47:39sparing nicely the hearts, lungs,
01:47:41and the spinal cord.
01:47:43You know, and I'll note
01:47:44again, kind of this awareness
01:47:46even a few years ago
01:47:47using FOLFOX based chemoradiation is
01:47:50really the appropriate treatment in
01:47:51this case.
01:47:53Great. So a post treatment
01:47:55PET CT showed a significant
01:47:57interval decrease in the hypermetabolic
01:47:59distal esophageal mass. The patient
01:48:01underwent a minimally,
01:48:02invasive distal esophageectomy with no
01:48:05complications
01:48:06and had a path complete
01:48:07response.
01:48:09So I'm hoping,
01:48:11maybe doctor Khan can comment
01:48:13a little bit on any
01:48:14surgical considerations you might think
01:48:16about, special surgical approaches or
01:48:18techniques.
01:48:20And then I'll I'll plant
01:48:21the seed for doctor Kourtmansky
01:48:22to comment on, would you
01:48:23give adjuvant treatment in this
01:48:25setting?
01:48:26If no,
01:48:27or and then maybe explain
01:48:28your rationale.
01:48:30Yeah. Thanks, Pam.
01:48:34Thanks, Pam. Yeah. So for,
01:48:36patients with g junction,
01:48:39tumors,
01:48:40there are multiple surgical approaches.
01:48:42The ones we use at
01:48:42at Yale are led by
01:48:44Dan Boffa, and he does
01:48:45a minimally invasive esophagectomy,
01:48:47where there's a the transthoracic
01:48:49component and an intra abdominal
01:48:51component as well too.
01:48:53Other approaches can be a
01:48:54transabdominal and cervical approach. Transheital
01:48:57esophagectomy is an approach that
01:48:58sometimes is used as well
01:49:00too.
01:49:01So there are multiple surgical,
01:49:02approaches towards this. But at
01:49:04Yale, we use a minimally
01:49:05invasive
01:49:06approach, through the chest and
01:49:08through the abdomen.
01:49:10Great. Thank you.
01:49:14So I so the first
01:49:15thing I would say is
01:49:16that patients that have a,
01:49:19pathologic complete response,
01:49:21are an excellent prognostic group.
01:49:23So your post treatment,
01:49:25staging,
01:49:26is relevant for prognosis.
01:49:30And so,
01:49:31median recurrence free survivals for
01:49:33these patients is can be
01:49:35measured as high as four
01:49:36to five years. So significantly
01:49:37better
01:49:38than patients who have any
01:49:40evidence of residual disease at
01:49:42the time of their surgery.
01:49:44And so that plays into
01:49:45the,
01:49:47the role of adjuvant
01:49:49immunotherapy,
01:49:50which was addressed in the
01:49:51CheckMate five seventy seven study,
01:49:53which was really only open
01:49:55to patients who had residual
01:49:57disease at the time of
01:49:58their surgery.
01:49:59It did show
01:50:00an improvement
01:50:02in
01:50:04recurrence free survival, about
01:50:06doubled that,
01:50:07that timing.
01:50:09And so we have no
01:50:10data to support the use
01:50:12of immunotherapy
01:50:13in this populate in the
01:50:14pathologic complete response patients, and
01:50:16so I would not offer
01:50:18any additional treatment at this
01:50:19time. Great. Thank you.
01:50:22Oh, yes. Reminder.
01:50:24So, doctor Raghav Sundar joins
01:50:26us next week. He will
01:50:28be helping to lead a
01:50:29new gastroesophageal
01:50:30cancer program here. So next
01:50:32year, we'll invite him back,
01:50:33to talk some about that.
01:50:35I'm not gonna go into
01:50:36the details of this, but
01:50:37this is just a reminder
01:50:38to me that we all
01:50:39need to be thinking about
01:50:40social drivers of health and
01:50:42health disparities as we care
01:50:43for all of our patients.
01:50:45Esophageal cancer is a male
01:50:48white predominant,
01:50:50cancer,
01:50:51and we are seeing more
01:50:52and more patients certainly in
01:50:54the state of Connecticut
01:50:56who are Asian, and it's
01:50:56also an Asian predominant cancer,
01:50:58especially,
01:50:59gastric cancer.
01:51:01So, I'll move on just
01:51:02to the next case for
01:51:03sake of time.
01:51:04Okay. So we are gonna
01:51:05talk a little bit about
01:51:06metastatic colorectal cancer, liver
01:51:09dominant, and we'll try to
01:51:10be speedy. We'll do it
01:51:11this in five minutes.
01:51:13So this is a forty
01:51:14nine year old with a
01:51:16history of polyps and normal
01:51:17colonoscopy
01:51:18presented with rectal bleeding.
01:51:20The patient has a history
01:51:21of depression and anxiety in
01:51:22addition to some other medical
01:51:24issues listed here.
01:51:25The patient has no history
01:51:27of alcohol use, a remote
01:51:29history of tobacco,
01:51:30smokes marijuana daily, and recently
01:51:32relocated from out of state.
01:51:34The patient's husband is disabled,
01:51:36and she serves as a
01:51:37primary caregiver.
01:51:39So let's start,
01:51:40with this with the initial
01:51:42workup here.
01:51:43And,
01:51:44I'm going to
01:51:46turn to my notes. Bear
01:51:47with me for one second.
01:51:49So,
01:51:51you guys can remind me
01:51:52since I can't find my
01:51:53notes. I think who did
01:51:54we decide was gonna talk
01:51:55about this one
01:51:57in terms of the initial
01:51:58workout?
01:51:59I'll move to this. Doctor
01:52:01Kormansky, I think I'll ask
01:52:02you.
01:52:03So I I mean, I
01:52:04have all the answers on
01:52:05the board. But,
01:52:07so
01:52:09but the most important thing,
01:52:11is getting the colonoscopy,
01:52:15and biopsy,
01:52:16which,
01:52:17confirms the diagnosis,
01:52:18but also,
01:52:20tells us the side of
01:52:21the tumor.
01:52:23So,
01:52:24left sided tumors,
01:52:25which is from the rectum
01:52:27up into the splenic flexure
01:52:28versus right sided tumors, which
01:52:30include the cecum,
01:52:32ascending colon, and transverse colon
01:52:34is all relevant,
01:52:35in terms of,
01:52:37our treatment decisions.
01:52:39As I mentioned earlier, we,
01:52:41prefer molecular testing on all
01:52:43patients, specifically
01:52:45microsatellite instability,
01:52:47and most importantly for metastatic
01:52:49patients, testing for KRAS,
01:52:51and BRAF mutations,
01:52:54staging studies,
01:52:56with a chest CT of
01:52:58the abdomen and pelvis.
01:53:01If there is liver metastases,
01:53:04consideration of an MRI,
01:53:06for better clarification,
01:53:08is important.
01:53:12And then pelvic MRIs for
01:53:13patients with rectal disease.
01:53:16Always multidisciplinary
01:53:18tumor board evaluation.
01:53:20I think it is very
01:53:21important to ask the question
01:53:23in somebody with metastatic disease.
01:53:26Is there a role for
01:53:28surgical therapy,
01:53:30of their metastatic disease or
01:53:31other local therapy?
01:53:33And it's important to ask
01:53:34that question
01:53:36early. They may not be
01:53:37relevant at the time, but
01:53:38at least you know what,
01:53:40what we're
01:53:42working towards. And there's a
01:53:43lot of patients that wind
01:53:45up,
01:53:46getting a lot of chemotherapy
01:53:48before referral for some sort
01:53:51of local therapy.
01:53:52Great. Thank you. I'm gonna
01:53:53pass to to doctor Baum
01:53:54just to,
01:53:56briefly comment on this patient
01:53:57certainly had some interesting,
01:53:59and challenging social drivers of
01:54:01health in addition to mental
01:54:03health challenges and is under
01:54:04the age of fifty.
01:54:06This is an area of
01:54:07expertise for Laura. So how
01:54:08how do you think about
01:54:09this when you have a
01:54:10patient in front of you?
01:54:11So I think when you're
01:54:12looking at a young person,
01:54:14and this is also my
01:54:15patient,
01:54:16who's,
01:54:17came here to be with
01:54:18her daughter, who's an undergrad
01:54:19at Yale. So she's an
01:54:20older, early onset patient. But
01:54:22I still think you're you're
01:54:23thinking,
01:54:25what are the are there
01:54:26family or fertility issues? Are
01:54:28there pragmatic issues? And are
01:54:30there medical issues related to
01:54:31the age? So I think
01:54:33for,
01:54:34all early onset patients, you
01:54:35do wanna make sure you've
01:54:37considered,
01:54:38you know, genetic and familial
01:54:39testing.
01:54:41You wanna consider fertility issues.
01:54:42And then you wanna consider
01:54:44pragmatic issues. You know, obviously,
01:54:46younger patients have childcare issues.
01:54:48They have they're more likely
01:54:49to be underinsured or uninsured,
01:54:51and they're more likely to
01:54:52have a school or career
01:54:54disruption.
01:54:55I also think that in,
01:54:57younger patients and particularly in
01:54:59people
01:55:00with a history of,
01:55:01prior mental health diagnoses or
01:55:03prior traumatic exposures,
01:55:06the cancer diagnosis,
01:55:08can be perceived as a
01:55:10as traumatic.
01:55:11And so the way we
01:55:12approach that from the beginning,
01:55:15makes a real difference in
01:55:17their experience.
01:55:19When we are provided when
01:55:21we when we experience trauma,
01:55:23our executive function,
01:55:25decompensates,
01:55:27and we lose sort of
01:55:28order and processing in our
01:55:30thoughts. And so helping
01:55:32the patient to identify
01:55:33those traumatic symptoms versus their,
01:55:36you know, cancer related symptoms
01:55:37can help them think more
01:55:38clearly,
01:55:40have better long term outcomes,
01:55:42improve their sense of order
01:55:44and their self observing capacity,
01:55:46and I think essentially help
01:55:48them feel,
01:55:50supported and that they're not
01:55:51alone,
01:55:52and that we're gonna take
01:55:53good care of them. So
01:55:54Thank you. Just to plug,
01:55:56we do have an early
01:55:57onset team here now. And
01:55:58then for all patients,
01:56:00regardless of age, we have
01:56:01a psycho oncology consult service
01:56:03and a and a pretty
01:56:04comprehensive social work,
01:56:06service.
01:56:07Great.
01:56:08So the, laboratory workup,
01:56:11CBC, CMP were normal, CA
01:56:13was elevated,
01:56:15colonoscopy
01:56:16showed a mass at ten
01:56:17centimeters,
01:56:18and path should moderate moderately
01:56:20differentiated adenocarcinoma,
01:56:21microsatellite stable, HER2 negative KRAS
01:56:24gene mutant.
01:56:26For sake of time, I'm
01:56:27gonna get into the management,
01:56:28so we're not gonna focus
01:56:29extensively on the imaging.
01:56:31So maybe I'll just move
01:56:32to the next slide.
01:56:34So this patient went on
01:56:35to receive FOLFOX and bevacizumab
01:56:37for nine cycles at an
01:56:38outside hospital. She moved to
01:56:40Connecticut, transferred her care to
01:56:42Smilo,
01:56:43reports increasing neuropathy,
01:56:45some additional symptoms,
01:56:46and interval CT showed stable
01:56:48disease in the liver and
01:56:49colon primary. She's presented to
01:56:51TumorBoard, which we have here
01:56:52essentially.
01:56:53Jill presented,
01:56:55interesting data on liver transplant
01:56:57and on thermal ablation. And
01:56:58maybe what we can do
01:56:59just in the limited time
01:57:00that we have is maybe
01:57:01doctors Khan and Du can
01:57:03talk about some of these
01:57:04additional
01:57:05liver directed therapies, including hepatic
01:57:07artery infusion, radiation, and perhaps
01:57:09thermal ablation.
01:57:12Great, Pam. So, so for
01:57:14patient with colorectal cancer liver
01:57:15metastases,
01:57:16you know, the staging workup,
01:57:18was great and nicely,
01:57:20presented. And for this patient
01:57:21in particular, the patient presented
01:57:23with bilobar hepatic metastases.
01:57:26Patients with bilobar hepatic metastases
01:57:28are not necessarily unresectable patients.
01:57:30They can be treated with
01:57:31potentially cured of liver resections
01:57:34with an ablation as well
01:57:35too can, also be incorporated
01:57:37into that, in the armamentarium
01:57:39as well.
01:57:40I do think it's important,
01:57:41to note that even though
01:57:43patients may present with unresectable
01:57:46hepatic metastases,
01:57:47they do,
01:57:49be seen by surgical oncologists,
01:57:52because there are patients with
01:57:53who initially present with unresectable
01:57:55liver metastases who may be
01:57:56resectable down the road, and
01:57:58resection of the primary and
01:57:59the metastatic tumor, might may
01:58:01provide cure. We have a,
01:58:03hepatic arterial infusion pump program
01:58:05right now that we have
01:58:06at Yale and,
01:58:07which is, going quite well.
01:58:09We've,
01:58:10operated on several patients with
01:58:11unresectable colorectal cancer liver metastases
01:58:14and unresectable intrahepatic cholangiocarcinoma.
01:58:18And, we are seeing good
01:58:19results and potentially patients that
01:58:21may become resectable or at
01:58:22least their liver disease is
01:58:24controlled also at this time
01:58:25frame. So there are different
01:58:26ways to treat this, and,
01:58:28I think it's important from
01:58:30a multidisciplinary
01:58:31approach to actually have the
01:58:32patient seen by a surgical
01:58:34oncologist early on if possible
01:58:36because
01:58:36sometimes too much chemotherapy
01:58:38may render someone unresectable down
01:58:40the road, and I think,
01:58:42that would be, one of
01:58:43the goals down the road.
01:58:44To you, Kevin.
01:58:47So the role of radiation
01:58:48for this metastatic rectal cancer
01:58:50patient is,
01:58:52debatable.
01:58:53However, I would say, in
01:58:55general, for rectal cancers and
01:58:58colorectal cancers,
01:58:59we do as, as, doctor
01:59:01Khan was just discussing, you
01:59:03know, consider that
01:59:04metastatic colorectal cancers are potentially
01:59:07curable.
01:59:09With more limited disease, I
01:59:10would say that
01:59:12the role of radiation
01:59:13is really in terms of
01:59:14local control in the pelvis
01:59:16and then also,
01:59:18the role of the question
01:59:19of whether SBRT may help
01:59:21in terms of clearing the
01:59:22liver.
01:59:23For this patient, there's, certainly
01:59:25enough bulky presacral lymphadenopathy.
01:59:28I would think that local
01:59:29recurrence even after an excellent,
01:59:30rectal surgery would still be
01:59:32a concern
01:59:33if you are going for
01:59:34cure to consider adding in
01:59:35pelvic there radiation for this
01:59:37patient.
01:59:41The different modalities
01:59:42for liver
01:59:44directed therapy are numerous and,
01:59:47poorly defined as to which,
01:59:50ones may, best benefit,
01:59:53the patient.
01:59:54You know, I'll just point
01:59:55out that,
01:59:58doctor Lacey brought up the
01:59:59collision,
02:00:00trial for colorectal cancers,
02:00:02looking at surgery versus,
02:00:04ablation for small tumors in
02:00:06the collision group.
02:00:07Following up on this is
02:00:08also enrolling currently on the
02:00:10collision x XL trial, the
02:00:13supersized version of collision,
02:00:15looking at,
02:00:16ablation versus SBRT in a
02:00:18randomized fashion, which is, really,
02:00:21I'm looking forward to as
02:00:22a radiation oncologist.
02:00:24Great. I think we're gonna
02:00:26do you wanna make one
02:00:26yeah. Gary gets the last
02:00:28word. I wanna
02:00:30actually, I just wanna make
02:00:31one quick comment about,
02:00:33you know, the choice of
02:00:34upfront chemotherapy for this patient.
02:00:36So she
02:00:37is,
02:00:38a young patient,
02:00:40with a left sided tumor.
02:00:41And we have,
02:00:43two two studies that have
02:00:45demonstrated that in combination with
02:00:47bevacizumab
02:00:48triplet therapy, so FOLFOXIRI
02:00:50as opposed to FOLFOX or
02:00:52FOLFIRI,
02:00:53do show demonstrations,
02:00:55an improvement in survival,
02:00:58as well as improvements in
02:01:00conversion to resection. So,
02:01:03I do think we should
02:01:04be thinking about which patients
02:01:06are candidates for triplet therapy,
02:01:09at the time of their
02:01:10diagnosis.
02:01:11Great. Well, we will end
02:01:13there. I wanna thank the
02:01:14panel for their comments and,
02:01:16really appreciate that. We think
02:01:17we have about a ten
02:01:18minute break. So thanks everybody.
02:01:25Now I know how Rodney
02:01:26Dangerfield felt. Come on.
02:01:30Okay. So this year, we
02:01:32had a very new format.
02:01:34I was I'm very honored.
02:01:36I should've said to the
02:01:37beginning to be working with
02:01:38co directors,
02:01:40Scott Huntington,
02:01:41Pam Koons,
02:01:43Kiran Turaga. But now, co
02:01:44director Sue Evans here is
02:01:46going to introduce a very
02:01:47special lecture because one of
02:01:48the things that's special about
02:01:49Yale and SMILE is our
02:01:51new technology,
02:01:52the innovative techniques we have,
02:01:54and who better to do
02:01:55it than one of our
02:01:55top radiation oncologists. So Sue.
02:02:04Thank you for that, Roy.
02:02:05Thank you, everyone, for settling
02:02:06back down. So lovely. So
02:02:08we are going to start
02:02:10off with,
02:02:12two presentations.
02:02:13One from doctor Udi Mendel
02:02:15about the spine oncology program
02:02:17at Yale, and then that
02:02:19will be followed by doctor
02:02:20Henry Park. Doctor Park is
02:02:22with radiation oncology, and doctor
02:02:24Mel Mandel is with neurosurgery.
02:02:27So please give them your
02:02:28attention and applause.
02:02:35Hey. Good morning, everybody.
02:02:38So I was asked to
02:02:39give a talk about, the
02:02:40spine program at Yale and
02:02:42specifically some of the stuff
02:02:43that's unique to SMILO,
02:02:46that we are doing here.
02:02:49So as you all know,
02:02:50we we walk
02:02:52I we all walk, very
02:02:54close with all of you.
02:02:55It's all about
02:02:57a multidisciplinary
02:02:59approach,
02:03:00where we deal with every
02:03:02patient very much in a
02:03:03individual's way.
02:03:05And, ultimately, over the last
02:03:06three years, we developed specific
02:03:08pathways within SMILO,
02:03:11to
02:03:13make
02:03:15treatment better for our patients.
02:03:16So we develop ERAS pathways
02:03:19for appropriate and urgent need
02:03:20for a quick biopsies and
02:03:21pathology on patients who come
02:03:23in with no cancer diagnosis,
02:03:24but very high suspicion for
02:03:26cancer.
02:03:27We develop pathways for a
02:03:29spinal cord
02:03:30spinal cord compression protocol. So
02:03:33patient coming into the emergency
02:03:34room, and they may have
02:03:36severe cord compression
02:03:38and rapid succession of workup
02:03:40needs to be done. So
02:03:41we now have an ERAS
02:03:43pathway for that that leads
02:03:45to emergent, urgent, or routine
02:03:47MRIs or urgent, emergent, routine
02:03:50CT monograms or biopsies. All
02:03:53of those things have been
02:03:54significantly improved over the last
02:03:57few years.
02:03:59Now why are we doing
02:04:00it? It's because spine tumors,
02:04:03is not unlike here, you
02:04:04get the specific talk about
02:04:06specific cancer,
02:04:07we're dealing with cancer all
02:04:08cancers. So spine tumors involve
02:04:10all cancer types. They can
02:04:12be
02:04:13aggressive,
02:04:14debilitating, causing significant pain.
02:04:17Some of these patient develop
02:04:19rapid development of neurological problems,
02:04:22as you all know.
02:04:23Patient can progress to paralysis,
02:04:26paralysis,
02:04:28bowel and bladder problems,
02:04:30and increase risk ultimately for
02:04:32with all the other things
02:04:34that goes along with these
02:04:35type of patient populations.
02:04:37What are the patients that
02:04:38we're talking about? Well, it's
02:04:39mostly the metastatic cancers that
02:04:40a lot of you dealing
02:04:41with, but we also have
02:04:42to deal with a lot
02:04:43of primary bone tumors. The
02:04:44chondrosarcomas,
02:04:45the sarcomas,
02:04:47are another separate patient population.
02:04:50And then we have the
02:04:51intradual, extradual tumors, whether it's
02:04:53a leptomeningeal
02:04:54disease, where we're relying on,
02:04:56no oncology colleagues,
02:04:58or whether it's the more
02:04:59of the benign patient populations.
02:05:01And so you can see
02:05:02when it comes to these
02:05:03spine tumors, they can be
02:05:05intradual, intramedullary,
02:05:07extradual, in the bones, all
02:05:09of those things. And each
02:05:10one of these patients really
02:05:11require
02:05:13a specific
02:05:14plan,
02:05:16for these particular patients.
02:05:18The majority of the patients
02:05:20we deal with are the
02:05:20metastatic tumors. So about one
02:05:22point two million new cancer
02:05:24cases diagnosed in the US,
02:05:27about more than half a
02:05:28million deaths a year.
02:05:30It's a major cause of
02:05:31death due to complications due
02:05:33to metastatic disease, and the
02:05:35skeletal system is the third
02:05:36most common site of metastases
02:05:38after lung and liver.
02:05:40The spine itself
02:05:41is the most common site
02:05:42of metastatic disease,
02:05:45and as many as ninety
02:05:46percent of patients with cancer
02:05:47will have spinal metastases
02:05:49at time of autopsy. So
02:05:51it's a huge patient population.
02:05:53And out of those, we
02:05:55don't see all of them,
02:05:56but up to thirty percent
02:05:57of these patients
02:05:58have some symptoms leading to
02:06:01us getting involved with this
02:06:03patient, whether it's pain or
02:06:04whether it's some neurological issues.
02:06:07The primary tumors are the
02:06:08ones I just mentioned. That's,
02:06:09again, a smaller group. It's
02:06:10forty fifteen to twenty percent
02:06:12of the patients we deal
02:06:13with.
02:06:15You guys all know about
02:06:16these require completely separate,
02:06:19different planning.
02:06:21Now why are these patient
02:06:22feel difficult
02:06:24from our point of view?
02:06:25Well, you guys don't know
02:06:26know all of this stuff.
02:06:27They are mostly elderly. They
02:06:29are immune deficient. They are
02:06:30on steroids.
02:06:31They have rapid weight loss.
02:06:33They already got some chemo
02:06:35radiation steroids, so healing issues
02:06:37for these wounds are critical.
02:06:39A lot of them are
02:06:40coagulopathic.
02:06:41A lot of them have
02:06:42wound issues,
02:06:43whether it's infection, whether it's
02:06:44wound breakdown.
02:06:46A lot of times, we
02:06:47don't know what we're dealing
02:06:48with. So establishing a diagnosis,
02:06:51operating on the right patient
02:06:53with the appropriate surgical plan,
02:06:57are all difficult.
02:07:00Patients,
02:07:01unlike most spine cans
02:07:03spine patients who get the
02:07:04routine degenerative work,
02:07:06in the degenerative ward, we
02:07:08like fusion. We like bone
02:07:09growth. We like healing.
02:07:11These patients rarely fuse,
02:07:14and so it leads to
02:07:16high rate of failure. And
02:07:17then timing of surgery, when
02:07:18is the right time to
02:07:19do something?
02:07:21So I'm gonna just skip
02:07:22through it just for the
02:07:23sake of time. But, again,
02:07:24a lot you know, immunocompromised,
02:07:26decreased white count. We see
02:07:28more and more patients who
02:07:29get cement injections. The multiple
02:07:31myeloma patients get a lot
02:07:32of those, cement injections for
02:07:34fractures.
02:07:35They get infected sometimes. Cement
02:07:37loves infection. It's very porous,
02:07:38very difficult to eradicate those
02:07:40infections from the cement.
02:07:42Steroids, lots of side effects
02:07:44related to these steroids.
02:07:46Coagulopathic,
02:07:47a lot of these patients,
02:07:48we wanna do surgery. We
02:07:49can't do surgery. A white
02:07:50count is high or low.
02:07:52The platelet count is very
02:07:53low.
02:07:55So they are at higher
02:07:56risk for DVTs.
02:07:58And then beyond all that,
02:08:00when we plan, you know,
02:08:01doing surgery, a lot of
02:08:02these tumors are very vascular.
02:08:03And so when you go
02:08:04in and you start removing
02:08:05this tumor,
02:08:06count on significant blood loss.
02:08:08And the renal cells, the
02:08:10liver cells, the the thyroid
02:08:12cell, the pheochromocytoma,
02:08:13all these tumors, and including
02:08:14the primary tumors,
02:08:16we anticipate
02:08:17liters of blood loss during
02:08:18these type of surgeries.
02:08:20And then,
02:08:22you know, when we do
02:08:22these embolization, there's always issues
02:08:24that are more specific for
02:08:25surgery. But if you guys
02:08:27remember for medical school, there's
02:08:28always issue with Adio Adamkiewicz.
02:08:30Adio Adamkiewicz is a is
02:08:31a blood vessel that feeds
02:08:33the spinal cord. And as
02:08:34they do these these angiograms
02:08:37and they're doing these embolization
02:08:38to cut on the blood
02:08:39loss, we get some information
02:08:41that sometimes we don't wanna
02:08:42know, and that's specifically about
02:08:43this artery of adamicoids that
02:08:45feeds the spinal cord. That's
02:08:47what we wanna see. This
02:08:48is a patient with a
02:08:48renal cancer, vertebral body. That's
02:08:51what it looks like before
02:08:52the embolizer. The whole vertebral
02:08:54body is one gigantic
02:08:56blood vessel con conglomerate. And
02:08:58so we need that preoperative
02:09:00planning,
02:09:01preoperative embolization,
02:09:04to cut down on the
02:09:05blood loss. And now we
02:09:06have a huge team at
02:09:08Smilow
02:09:08that actually available twenty four
02:09:10seven to give us, this
02:09:12particular
02:09:13image and then make it
02:09:14look like this. So after
02:09:16the embolization, they can go
02:09:17in, they can put cores,
02:09:18glue, whatever it is, in
02:09:20order for us to be
02:09:21able to attack these,
02:09:24vertebral bodies and cut on
02:09:25the blood loss during the
02:09:26surgery.
02:09:28Worn issues is always a
02:09:29big deal, as I mentioned.
02:09:30So collaboration with the plastic
02:09:32surgeons is very critical,
02:09:35on a lot of these
02:09:36cases that we do. So
02:09:37we have a close collaboration
02:09:38with our plastic service, and
02:09:39now they have a team
02:09:41that's actually available for wound
02:09:43closures in patients with cancer.
02:09:45Ultimately, when it comes to
02:09:46primary versus metastatic,
02:09:49different goals, different adjuvant therapy,
02:09:51different surgical
02:09:52technique
02:09:54with goal for primary is
02:09:56to cure patients of cancer.
02:09:57So a lot of the
02:09:58cases that we do
02:10:00with the primary tumor, the
02:10:02massive surgeries,
02:10:04but the idea is to
02:10:05cure them off the cancer.
02:10:06If we can get clean
02:10:07margins around these,
02:10:09we can cure the patients
02:10:10of a cancer versus metastatic.
02:10:11A lot of these are
02:10:12palliation.
02:10:13Adjuvant therapy for primary, very
02:10:14limited, metastatic available, and the
02:10:16surgical technique is very different.
02:10:18The big issue with spine
02:10:19tumors
02:10:20is what I talk to
02:10:22my fellows, colleagues, residents fellow,
02:10:25is that sometimes you find
02:10:26yourself
02:10:27doing the wrong operation on
02:10:29the wrong patient by the
02:10:30wrong surgeon.
02:10:31And many times, what happen
02:10:32is you get the emergency
02:10:35room spine person on call,
02:10:37and they apply degenerative trauma
02:10:39principles,
02:10:41for patients who have spine
02:10:42cancer, and those things just
02:10:44many times
02:10:46don't go
02:10:47together.
02:10:48So a couple of cases
02:10:49here. This is a twenty
02:10:51three year old lady.
02:10:52She fell in the office,
02:10:55no cancer diagnosis,
02:10:57start having back pain, and
02:10:59neurologically
02:11:00moving everything, just having back
02:11:01pain. This is what the
02:11:02MRI looks like when she
02:11:03comes to the emergency room
02:11:04complaining of no onset back
02:11:06pain after she fell. You
02:11:08can see the spinal cord
02:11:09is squashed.
02:11:10Something is going on. We
02:11:12have no cancer diagnosis, but
02:11:14the patient's moving her legs.
02:11:15Just a lot of pain.
02:11:16Why is it hurt?
02:11:17Because the vertebrae is broken
02:11:19and because there is significant
02:11:20compression on the dura. The
02:11:22dura is very painful. When
02:11:24something compress the dura, there's
02:11:25a lot of pain related
02:11:26to it. Well, what is
02:11:27it?
02:11:28Don't know. She's in the
02:11:29emergency room. They consult the
02:11:32spine to an on call.
02:11:34The spine surgeon comes in,
02:11:35examine. The patient's moving everything.
02:11:38I don't like the fact
02:11:39that the spinal cord is
02:11:40severely compromised. I'm worried about
02:11:42it. Here's the CT scan.
02:11:45And the the surgeon take
02:11:46the patient to surgery
02:11:48and does a decompression, laminectomy.
02:11:50Right? That's what it looks
02:11:51like after the surgery. The
02:11:52back of the spine has
02:11:53been removed. That's what it
02:11:55looks like after the surgery.
02:11:56And you can see the
02:11:57MRI on the left. About
02:11:59ten percent of the tumor
02:12:00has been removed. The cord
02:12:01looks better, happier,
02:12:03but all this stuff around
02:12:04the vertebral body is still
02:12:05there.
02:12:07The biopsy
02:12:08intraoperative
02:12:09specimen comes back as lymphoma
02:12:11on this patient. This patient
02:12:13cannot get radiation, cannot get
02:12:14a lot of the adjuvant
02:12:15therapy that you guys wanna
02:12:16do on her, and within
02:12:18that's delayed the treatment options
02:12:21for her. The wound got
02:12:22infected,
02:12:24two weeks after the surgery.
02:12:25Could not get any radiation
02:12:27treatment. The patient died in
02:12:28three months after the after
02:12:30the surgery.
02:12:33I'm gonna skip that because
02:12:35that's more
02:12:37specific. There is another case
02:12:39right here. This is a
02:12:40patient who come into the
02:12:41emergency room complaining of back
02:12:43pain, bowel, bladder incontinence over
02:12:45twenty four hours, and foot
02:12:47drop. And the MRI shows
02:12:49a sacral lesion.
02:12:50The sacrum has a lot
02:12:51of tumor. The canal is
02:12:52completely full of tumor. The
02:12:54s one air roots are
02:12:55completely displaced to the side.
02:12:57No diagnosis,
02:12:59but
02:13:00neurological
02:13:01deficits.
02:13:02Balm bladder incontinence,
02:13:04back pain, foot drop.
02:13:06What is
02:13:07it? Take care of the
02:13:08take the patient to surgery,
02:13:09not take pay the patient
02:13:10to surgery, nose surgery, spine
02:13:12surgery that's called,
02:13:14and they do a laminectomy,
02:13:16they decompress the nerve roots,
02:13:18the patient is doing great
02:13:19for six months,
02:13:21Bone bladder recovered.
02:13:24Foot weakness
02:13:25recovered.
02:13:26And, pain has gotten better.
02:13:28And this patient presents six
02:13:30months later
02:13:31from that hospital to our
02:13:33hospital, and he says, hey,
02:13:34doctor Mendel. I think one
02:13:36of the screws that he
02:13:36put in there is poking
02:13:37up through my back,
02:13:39and he points to this.
02:13:41That's what his back looks
02:13:42like. You can see the
02:13:42incision in the back. And
02:13:44he points this. He said,
02:13:45I think that's a screw.
02:13:46One of those screws are
02:13:47coming out, getting loose.
02:13:49That's what the MRI looks
02:13:50like, and you can see
02:13:51what it looks like now.
02:13:53This was a chondrosarcoma
02:13:55that had an intralesional
02:13:56resection.
02:13:57And within six months, this
02:13:59tumor is now filling up
02:14:00the whole canal.
02:14:03And the, quote, screw that's
02:14:04loose is actually that bump.
02:14:06It's a tumor recurrence to
02:14:08the path of the surgery.
02:14:10These patients should have had
02:14:12a completely different surgery, considering
02:14:14an unblocked resection with clean
02:14:15margins and not going in
02:14:17there, and decompress the nerve
02:14:18roots,
02:14:19which gives temporary relief for
02:14:21about six months, and that's
02:14:22it.
02:14:23This is a patient who
02:14:24potentially had a chance for
02:14:25a cure. This patient now
02:14:27has a death certificate on
02:14:28his name because this patient
02:14:30is gonna ultimately
02:14:31die from this. So the
02:14:32decision making of making
02:14:34a the right choice
02:14:36is super critical. This is
02:14:37the patient.
02:14:39Not only that, he also
02:14:40had a broken rods from
02:14:41the surgery. It was done
02:14:42at this other hospital. There's
02:14:44no fusion.
02:14:45Some of the pain is
02:14:46coming from the tumor growth.
02:14:47Some of the pain is
02:14:48coming from the rot broken.
02:14:50Took him to surgery.
02:14:52We did an unblocker section
02:14:53for that little bump that
02:14:55was sticking out of his
02:14:56back, and you can see
02:14:57the plastic surgeons then had
02:14:59to put it all together
02:15:00for us to remove
02:15:01this specimen.
02:15:03Here is a woman with
02:15:04breast cancer
02:15:05presenting to her oncologist with
02:15:07the fractures of two vertebrae
02:15:09in the back.
02:15:12And so there's by there's
02:15:14mechanical instability for the spine
02:15:16significant. She's presenting with significant
02:15:19amount of neck pain.
02:15:21C seven t one are
02:15:22broken. They are a little
02:15:23bit broken, not severely, but
02:15:25they are broken. They are
02:15:26pushing a little bit on
02:15:26the spinal cord.
02:15:28This patient, instead of referred
02:15:30or being presented or discussed
02:15:32surgical options here, ended up
02:15:34getting radiation, and that's what
02:15:35it looks like a month
02:15:37after the radiation. You can
02:15:38see and I I don't
02:15:40wanna blame my radiation oncologist,
02:15:41but you can see
02:15:43that what happened. Right? The
02:15:44two vertebraes are not able
02:15:45to hold this. Right? Completely
02:15:47collapsed. The tumor shrunk down,
02:15:48but the vertebraes collapsed.
02:15:50Now the spine is falling
02:15:51over, and there's severe spinal
02:15:53cord compression.
02:15:55This patient should not have
02:15:56had
02:15:57radiation, should have had mechanical
02:15:59restoration or mechanical stability, and
02:16:01then followed radiation down the
02:16:02road. Look at the CT
02:16:04scan. There's no no anterior
02:16:05column support.
02:16:07The patient is in agonizing
02:16:09neck pain, walking to the
02:16:10office, holding her neck like
02:16:11this, cannot move.
02:16:13So we ended up doing
02:16:15this surgery. You can see
02:16:16on the, on the right
02:16:17side,
02:16:18you know, a lot of
02:16:19reconstruction
02:16:21with front back surgery. And
02:16:22that's what it looks like
02:16:23at the end, when the
02:16:24front and the back has
02:16:25been reconstructed.
02:16:27Again, was it the right
02:16:28choice?
02:16:29That's where every patient has
02:16:31a specific
02:16:33need for a for a
02:16:35multidisciplinary
02:16:36input from all services. This
02:16:37is a patient with melanoma,
02:16:39pathologic fracture of the vertebra.
02:16:42There is severe cold compression.
02:16:44Patient presenting with,
02:16:46back pain,
02:16:48seen at an outside facility.
02:16:50And how did they treat
02:16:51the back pain? Well, a
02:16:52CAT scan shows that the
02:16:53bone is eaten away. The
02:16:55back wall is missing,
02:16:56and this patient underwent the
02:16:58fibroplasty.
02:16:59Why? To help with
02:17:01the back pain because the
02:17:02patient has a pathologic fracture.
02:17:03You can see the cement
02:17:04here.
02:17:06Okay. Here is the an
02:17:07x-ray, and look at the
02:17:08bottom picture. Where else is
02:17:09the cement? It's not just
02:17:10in the vertebrae.
02:17:11It's in the spinal canal.
02:17:13The cement leaked through the
02:17:14back of the wall
02:17:16into the spinal canal compressing
02:17:18the spinal cord.
02:17:19So there are multitude of
02:17:20issues here. Right? And that
02:17:22you can see right here
02:17:23on the CAT scan,
02:17:24the cement had leaked to
02:17:25the back into the spinal
02:17:26canal.
02:17:27And the patient is doing
02:17:28fine. The back pain got
02:17:29better. No neurological deficits likely
02:17:32because there's not a lot
02:17:33of cement in the canal.
02:17:34But that was not the
02:17:35right choice for this patient
02:17:36to get through cement injection
02:17:38to help with the back
02:17:39pain. Did the back pain
02:17:40got better? Yes. For about
02:17:42three weeks, only for it
02:17:43to break further
02:17:44and, you know, ultimately require
02:17:46an anterior transthoracic
02:17:47approach,
02:17:49you know, and, and fixing
02:17:51all this. So I always
02:17:52say, you know, we tend
02:17:53to think like spine surgeons,
02:17:55and we should think like
02:17:56spine surgeons on a lot
02:17:57of the trauma
02:17:59injuries, a lot of the
02:18:00infection issues. But when you're
02:18:01dealing with cancer, you gotta
02:18:03think as a cancer person
02:18:05first,
02:18:06then think about it as
02:18:07a spine surgeon. It should
02:18:09be reversed. And it's very
02:18:10hard for spine surgeons
02:18:12to think like that. We
02:18:12are very focused on neurological
02:18:14deficits. We are very focused
02:18:15on fixing it,
02:18:17and the fix may be
02:18:18a short fix, but not
02:18:20a long term fix. So
02:18:22at SMILO, we develop a
02:18:23huge multidisciplinary
02:18:25team
02:18:25where we have a high
02:18:27level of collaboration, not amongst
02:18:29just medical oncology, radiation oncology,
02:18:31and doctor Parker is gonna
02:18:32talk about it, but also
02:18:34about, you know, collaboration with
02:18:37our colleagues
02:18:39in surgery.
02:18:40This is a lady with
02:18:41thyroid cancer. You can see
02:18:42the vertebrae is broken in
02:18:43the front. She has a
02:18:45c six or c five
02:18:46fracture.
02:18:48The cord is is,
02:18:49severely compromised, with with cord
02:18:51compression.
02:18:53There's a CAT scan,
02:18:55and, you know, I was
02:18:56involved with developing this thing
02:18:57called the spine instability in
02:18:59your plastic scale. A lot
02:19:00of you guys calling and
02:19:01you say, hey. Is that
02:19:02spine stable? Is it gonna
02:19:04break more? Is it not
02:19:05gonna break more? Can we
02:19:06leave it like that? We
02:19:08start the chemo. Can we
02:19:09do the radiation?
02:19:10So we developed this thing
02:19:11called spine instability in your
02:19:12plastic scale. It meant for
02:19:14you, not for me,
02:19:16for many of the medical
02:19:17oncologists. A lot of the
02:19:18radiation oncologists will call and
02:19:19say, hey. We're getting ready
02:19:20to radiate. Is is this
02:19:21spine stable?
02:19:23We have a scale. We
02:19:24add the points.
02:19:25There are a lot of
02:19:26surgeons, radiation oncologists, and medical
02:19:29oncologists.
02:19:30We all got together, and
02:19:31we developed it over a
02:19:32two year course, and we
02:19:33developed this,
02:19:35this thing. And if you
02:19:36add your points, then you
02:19:37can figure out whether the
02:19:38spine is stable, zero to
02:19:40six points, potentially unstable or
02:19:41unstable.
02:19:43This patient, if you add
02:19:44the points, were unstable,
02:19:46had an angiogram,
02:19:47Christmas tree lighting, very vascular
02:19:50tumor,
02:19:50and you can see the
02:19:52plan. So we're now developing
02:19:53very nice reconstruction
02:19:55vertebraes for a lot of
02:19:56these patients to make it
02:19:57patient specific.
02:19:59We used to put a
02:20:00piece of bone in there,
02:20:01a piece of fibula, iliac
02:20:02crest. We don't do this
02:20:03anymore. ENT gets us the
02:20:05exposure.
02:20:06This is during surgery. This
02:20:08is what it looks like
02:20:09after the surgery. The whole
02:20:10vertebra is removed, and we're
02:20:12able to restore mechanical stability.
02:20:14This patient doesn't need to
02:20:15wear a color, and she
02:20:16can move on with, yeah,
02:20:17everything else. You can see
02:20:18their vertebra has been removed.
02:20:20The spine has been already
02:20:21reconstructed.
02:20:22And then we follow with
02:20:23radiosurgery,
02:20:24and we're gonna talk about
02:20:25it very quickly. I'm gonna
02:20:26be done in five minutes.
02:20:28Is that okay?
02:20:30Yeah. We'll just take it
02:20:31out of lunch. No problem.
02:20:33We'll be good.
02:20:34So we're gonna talk about
02:20:35the radius surgery here. I
02:20:37wanna talk about this case.
02:20:38This is a case against
02:20:39Mosmilo. This is a thirty
02:20:40four year old with a
02:20:41giant cell tumor that destroyed
02:20:43the whole pelvis. There's nothing
02:20:45left of the pelvis here.
02:20:49And that's been growing for
02:20:50many, many, many years. It's
02:20:51not to the point that
02:20:52the whole pelvis is one
02:20:54gigantic tumor.
02:20:55What do you do for
02:20:56that? Right? I mean, it
02:20:58used to be that's it.
02:20:58That's the end of the
02:20:59world. But this is a
02:21:00giant cell tumor. It's nowhere
02:21:02else other than here.
02:21:04Right? It's not metastasized
02:21:06yet. And but is there
02:21:07any surgical
02:21:09options for these type of
02:21:10patients?
02:21:11Well,
02:21:12look at this. Right? It's
02:21:14coming. It's just filling up
02:21:15the whole pelvis. How do
02:21:16you fix something like that?
02:21:19Not only that. Why is
02:21:20she at Smilow? Pain.
02:21:22Severe amount of pain. This
02:21:24patient is unable to get
02:21:25out of bed, already paralyzed,
02:21:28or kind of moving her
02:21:29legs barely, just her feet.
02:21:31Look at the amount of
02:21:31pain medications, including
02:21:34sixteen hundred microgram of ketamine
02:21:35infusion at thirty milligrams an
02:21:37hour.
02:21:38They couldn't
02:21:39nothing could help with the
02:21:40pain management for this patient
02:21:42other than this. And the
02:21:44idea,
02:21:45what are you gonna do
02:21:46here? And
02:21:47we
02:21:48brought the option of doing
02:21:49something that sounds nutty in
02:21:52hemicoorectomy.
02:21:53Have you guys ever managed
02:21:54these type of patients?
02:21:55This is this is the
02:21:56kind of cases where you
02:21:58literally cut the body in
02:21:59half.
02:22:00And is it being done?
02:22:01Yes. I mean, there are
02:22:03many many studies out there
02:22:04that show
02:22:05that this can be done.
02:22:07I was involved
02:22:08in one of these cases.
02:22:11A lot of surgeons are
02:22:12involved with with these when
02:22:13it comes to need for
02:22:14ligation of the aorta and
02:22:15the vena cava and so
02:22:16on and so forth. The
02:22:17idea was about pain management
02:22:20for this patient. Is there
02:22:21any other option? The pain
02:22:22gets better after this type
02:22:23of crazy surgeries
02:22:25and a lot of plastics
02:22:26involved.
02:22:27This is the case that
02:22:28was involved with. This is
02:22:29the picture from the operating
02:22:31room, and you can see
02:22:32how many surgeons are involved.
02:22:34The list is on the
02:22:35left.
02:22:36These are massive two to
02:22:38four step surgeries. I'm taking
02:22:39it to the extreme here,
02:22:40but that's the type of
02:22:41situation that a multidisciplinary
02:22:43component is very critical.
02:22:45And, ultimately, that's what it
02:22:47looks like. This is a
02:22:47picture from this case. When
02:22:49these cases are over, the
02:22:50plastic guys then needs to
02:22:51kinda put it together. What
02:22:53did we ended up doing?
02:22:54This patient was not a
02:22:55candidate for this type of
02:22:56surgery, and we can do
02:22:58it at Smilow. So I'm
02:22:59just showing that these these
02:23:01are options that are available.
02:23:03And, you know, we now
02:23:05offered
02:23:06there was not an option
02:23:07when it comes to the
02:23:08pain control. We offered this
02:23:09thing called cingulotomy,
02:23:10where our functional neurosurgery colleagues
02:23:13were able to do a
02:23:13cingulotomy.
02:23:15And what a cingulotomy, in
02:23:16essence, is a procedure where
02:23:17you do a lesion in
02:23:18the brain,
02:23:20and the patient's
02:23:21pain perception goes away.
02:23:23So,
02:23:25these are known options for
02:23:27these type of patients.
02:23:29We ended up doing a
02:23:29singularity on this patient. This
02:23:31case was presented in multiple
02:23:33tumor board. And after multitude
02:23:35discussion with all the teams
02:23:36involved,
02:23:37we have elected to do
02:23:38this, and you can see
02:23:40how we choose the spots
02:23:41in the brain to go
02:23:42ahead and do these lesions.
02:23:44And you can see,
02:23:46following this, the patient within
02:23:47two weeks was discharged on
02:23:48fentanyl three hundred microgram patch
02:23:50to the skin,
02:23:51with no pain perception.
02:23:54So
02:23:56in that sense, I'm gonna
02:23:57stop. I can show you
02:23:58many other, treatment options that
02:24:00we are having available. But
02:24:01ultimately, for the sake of
02:24:02time,
02:24:03we're gonna talk about our
02:24:04radiosurgery
02:24:05program that we have developed,
02:24:06at SMILO. We can't do
02:24:08spine surgery without having a
02:24:10very robust radiation
02:24:12oncology treatment, especially with the
02:24:14stereotactic radiosurgery program.
02:24:16We do have, for sake
02:24:17of publicity, we do have
02:24:18a spine tumor board
02:24:20that,
02:24:22Jen, have helped us establish.
02:24:24I can see Jen here,
02:24:25but she's helped us create.
02:24:26And so anybody who wants
02:24:28to have, like, patients, wanna
02:24:29discuss it, we have a
02:24:30weekly spine tumor board that
02:24:32all of the team gets
02:24:33together. Each medical oncologist, radiation
02:24:35oncologist could get a Zoom
02:24:36link, present the case, and
02:24:38then in a multidisciplinary
02:24:39fashion, we can discuss those
02:24:41type of cases.
02:24:42So with that sense, I'm
02:24:43gonna let, Park keep going
02:24:44with this. Thank you so
02:24:46much, doctor Mandel. That was
02:24:47fantastic.
02:24:52Well, welcome, doctor Park. And,
02:24:54doctor Park, our job in
02:24:55radiation oncology is to take
02:24:57care of residual disease. I'd
02:24:59like you to take care
02:25:00of some lost minutes, please.
02:25:04Let's see this this one.
02:25:06It's you download that. Right?
02:25:11It's are you That one.
02:25:12Yeah. Park. Yeah. That one.
02:25:14There we go. Good.
02:25:16Alright. Thanks very much for
02:25:17the invitation today. I've I've
02:25:19I've
02:25:20it was it was really
02:25:21great segue to talk about,
02:25:22the the the new tools
02:25:23we have in radiation therapy
02:25:24as well. I'm gonna talk
02:25:25about two things, mostly focus
02:25:27on the first one that
02:25:29we have already, then the
02:25:30second one we're hoping to
02:25:31get soon in in in,
02:25:32in twenty twenty six, proton
02:25:34therapy. But I'll first focus
02:25:36on the pet guided radiotherapy,
02:25:38my disclosures.
02:25:39Sextrotal beam radiation is really
02:25:41typically different by it it
02:25:42it's usually given by, from
02:25:43from from a linear accelerator,
02:25:45produces high energy X rays
02:25:47to treat cancer.
02:25:48So for each patient, we
02:25:49typically customize
02:25:51the the delivery and and
02:25:52of radiation
02:25:53using IMRT or SBRT, which
02:25:55you've, I'm sure, heard of
02:25:56before. This is guided by
02:25:57daily imaging before each treatment.
02:25:59However, some tumors do move
02:26:01as the, as the patient's
02:26:03breathe, especially in the lung
02:26:04and the GI tract.
02:26:07FLoNNX one is a new
02:26:08machine we have that we
02:26:09installed last year. It's a
02:26:11novel radiation technology that has
02:26:13the capability of using the
02:26:14tumor's own PET signal itself
02:26:16as a fiducial, meaning as
02:26:17a marker, to be able
02:26:18to track the radiation beams
02:26:19in real time. So So
02:26:21normally, we do what's called
02:26:22a four d scan. We
02:26:23can track a lung tumor,
02:26:24for example, as you breathe,
02:26:25and we target the whole
02:26:26area where it goes in
02:26:28real time. But what what
02:26:29we can do here, though,
02:26:30is is as a tumor
02:26:31moves, the radiation will actually
02:26:32follow along the tumor so
02:26:34we don't have to treat
02:26:35that whole area.
02:26:36So this is called biology
02:26:38guided
02:26:39for BGRT.
02:26:41It's not for a sixth
02:26:42therapy.
02:26:44So now the FDA has
02:26:45approved this for lung and
02:26:47bone only. So it's for
02:26:48both primary and secondary cancers.
02:26:49So this really links nicely
02:26:51with the METs that we
02:26:52just talked about in the
02:26:52bone. It can be used
02:26:53there as well. Mostly, we've
02:26:54been using it for lung
02:26:55so far. It could be
02:26:56delivered with s b r
02:26:57so really anything that could
02:26:58be otherwise s b r
02:27:00t could be considered for
02:27:01this as well. You can
02:27:03also use this machine that'll
02:27:04treat with IMRT and SBRT
02:27:05as as we're doing right
02:27:06now. For most of the
02:27:07machine right now is being
02:27:08used for for the most
02:27:09machine time right now is
02:27:10being used for that,
02:27:12purpose.
02:27:13So the biological advantages of
02:27:15Syntyxt therapy. So let's say
02:27:16we have, this is a
02:27:19a simplified view over all
02:27:21of this, but this is,
02:27:22what we call GTV or
02:27:23the gross disease is a
02:27:24tumor. We add some margin
02:27:26to it called PTV.
02:27:27So what we do with
02:27:28SBRT, again, that GTV will
02:27:30move, and, and then we'll
02:27:32put some margin around that.
02:27:33So to capture all of
02:27:34that becomes the the actual
02:27:36target that we end up
02:27:37treating.
02:27:38Now with BGRT, we can
02:27:39track the PET signal in
02:27:41real time to decrease the
02:27:42size of the radiation field
02:27:43for where the high dose
02:27:44goes. We can increase our
02:27:46confidence in fully hitting the
02:27:47target because sometimes the tumors
02:27:49move,
02:27:50the the the they're they're
02:27:51it's possible to have it
02:27:52move in different ways than
02:27:53it was at the simulation
02:27:54scan. So and and we're
02:27:56seeing that now that we're
02:27:56use we're using this technology.
02:27:58It's actually pretty disconcerting sometimes
02:28:00to see that, that that
02:28:01what we what we think
02:28:02it's it's happening is actually
02:28:03maybe not happening quite as
02:28:05as precisely as we as
02:28:06we thought, for for, for
02:28:09how long tumors move.
02:28:10So and then, the other
02:28:11thing is that we can
02:28:12decrease the risk of side
02:28:13effects potentially, although that hasn't
02:28:15yet been proven in clinical
02:28:16trials.
02:28:18So the installation happened here,
02:28:20in April two thousand twenty
02:28:21three.
02:28:22We installed the world's fifth
02:28:24machine of this type. We
02:28:26were the first in the
02:28:27East Coast. They're now six
02:28:28total overall.
02:28:29Now we've installed the Syntex
02:28:31therapy part in December. So
02:28:32a little about, it looks
02:28:34like about nine months ago
02:28:35now, soon after the FDA
02:28:36approval for lung and bone
02:28:38tumors.
02:28:39Here was the ribbon cutting,
02:28:40and we did our, which
02:28:41we did in January and
02:28:42then the first treatment in
02:28:43March.
02:28:45So Syntex therapy, the the
02:28:47main, patient selection criteria, we
02:28:48look for a target that's
02:28:49about two to five centimeters.
02:28:50And you all don't have
02:28:51to worry about this too
02:28:52much. We'll figure this out
02:28:53when you send patients to
02:28:54us for SBRT.
02:28:55And, basically, if you have
02:28:56lung or bone tumors, that
02:28:58could be SBRT, just just
02:28:59ask to see if this
02:29:00is, something that the patient
02:29:01might be eligible for. The
02:29:03main thing you all have
02:29:04to know is that we
02:29:05do need a PET scan
02:29:06beforehand to see if the
02:29:07SUV is high enough. So
02:29:08just think of the number
02:29:09six, SUV of six being
02:29:10being the the the main
02:29:11cutoff there. So if it's
02:29:13hot enough, then they may
02:29:14be a candidate for this.
02:29:16So lung or bone tumor
02:29:17of any type with an
02:29:18SUV of six.
02:29:20And then, the signal has
02:29:21to be a little bit
02:29:22homogeneous as well throughout the
02:29:23target.
02:29:24So the basic procedure is
02:29:25is similar to what we
02:29:27normally do. We do a
02:29:28CAT scan simulation,
02:29:29but the one extra thing
02:29:30we do is a PET
02:29:31modeling scan afterwards. So usually
02:29:33a couple days later, we'll
02:29:34do a PET modeling where
02:29:35we inject the FTG, a
02:29:36little higher dose than you
02:29:37would use for a diagnostic
02:29:38PET scan, the fifteen millicuries.
02:29:40And then at the PET
02:29:40modeling, we, we actually basically
02:29:42run a PET scan in
02:29:43our department of radiation oncology
02:29:45on the treatment machine itself
02:29:47in the treatment position so
02:29:48that we can see exactly
02:29:49that the target is is
02:29:51con is confirmed to be
02:29:52within the expected bounds, and
02:29:53we confirm that it passes
02:29:55the the basically, the tests
02:29:56it needs to do in
02:29:57order to, proceed with treatment
02:29:59planning.
02:29:59So then we can actually
02:30:01run it through the system
02:30:02even based on a diagnostic
02:30:03pet beforehand just to make
02:30:04sure that it's hot enough
02:30:05and big enough to to
02:30:06treat. And then we do
02:30:07similar preparations for each fraction.
02:30:09So if we do three
02:30:10fractions of of of BGRT,
02:30:12we'll do the,
02:30:13this the PET modeling and
02:30:15then three, then then then
02:30:16we do the FDG injection
02:30:17each of the fractions as
02:30:18well. So you might, you
02:30:19might end up with four
02:30:20injections in total.
02:30:23So I'll just give a
02:30:24couple cases here. Case one
02:30:25is a, the the first
02:30:26successful case we did was
02:30:27a left lung metastasis from
02:30:29endometrial adenocarcinoma
02:30:31about two point eight centimeters,
02:30:32somewhat close to the chest
02:30:33wall and somewhat close also
02:30:35to the tracheal bronchial tree,
02:30:36which are are two organs
02:30:38that we we watch out
02:30:38for with SVR team. SUV
02:30:40was nine point five.
02:30:42So past the evaluation that
02:30:43we did on a diagnostic
02:30:44PET, it passed the modeling
02:30:45scan that we did on
02:30:46the patient. We ended up
02:30:47treating sixty gray and five
02:30:48fractions with the, you know,
02:30:49here's the sixty gray line
02:30:51is right here, where the
02:30:52target is, and then forty
02:30:53gray and twenty gray. So
02:30:54this is a kind of
02:30:55a normal scatter that we
02:30:56expect. And then you can
02:30:57see here on the treatment
02:30:58machine, we're getting the PET
02:30:59signal that we're tracking as
02:31:00well.
02:31:01And here's, I'm not gonna
02:31:02go through the full details
02:31:03of this. But, basically, this
02:31:04this I meant to point
02:31:05out that we can see
02:31:06exactly what we're targeting now.
02:31:08We we have the PET
02:31:09signal each day. We we
02:31:11have that mashed up to
02:31:12the PET signal we get
02:31:13at the modeling, and then
02:31:14we can actually see where
02:31:15the dose actually went afterwards,
02:31:17which is something that we
02:31:18we model with with normal
02:31:19radiation therapy, but can't actually
02:31:21tell for sure at the
02:31:22end exactly where the dose
02:31:23went. But here here we
02:31:24do. We can see that.
02:31:26Now this second case, was,
02:31:28was from from from from,
02:31:30is is also right lung
02:31:31metastasis.
02:31:32There's also while we were
02:31:33working this patient up with
02:31:34a CT with contrast, we
02:31:35also found a left kidney
02:31:36lesion that was new compared
02:31:37to the PET scan the
02:31:38patient had three months prior.
02:31:40So after speaking to the
02:31:40medical oncology at Dana Farber,
02:31:42who who could send the
02:31:43patient to us, for this,
02:31:44then we ended up getting,
02:31:46we ended up deciding to
02:31:47treat both. Now we had
02:31:48just gotten an upgrade at
02:31:49that point of the software
02:31:50that let us treat both
02:31:51at the same time in
02:31:53the same plan. So, So,
02:31:54we were actually the first
02:31:55in the world to ever
02:31:55do this. So, this passed
02:31:58the TED evaluation, and we
02:31:59ended up treating forty gram
02:32:00five fractions to both, and,
02:32:02and and we we successfully
02:32:03completed the first first of
02:32:05these hybrid Syntex and SBRT
02:32:07plans at at once, which
02:32:09is remarkable mostly in the
02:32:10fact that not that we
02:32:12could treat two things, that
02:32:13that's,
02:32:14in and of itself not
02:32:15that interesting because we could
02:32:17have done that separately. But
02:32:18the main idea is if
02:32:19you have two or three
02:32:21or eventually five or ten
02:32:22lesions in the lung, for
02:32:23example, and you're gonna be
02:32:24treating
02:32:25all at once,
02:32:26that is something you could
02:32:27plan together. And then that's
02:32:28a future that we can
02:32:29imagine, could be really exciting
02:32:31that we can treat, not
02:32:32only oligometastatic disease, but potentially
02:32:34polymenastatic disease in the future.
02:32:36But right now, we're not
02:32:38at that point yet. We
02:32:39only can do one
02:32:40site with the BGRT,
02:32:42but now we can combine
02:32:43it with SBRT to other
02:32:44sites as well in a
02:32:45single plan. So I'm not
02:32:47gonna go through this much
02:32:48more. So the base of
02:32:48our vision for the future
02:32:49with Syntex at Yale, we're
02:32:50hoping to and then overall,
02:32:52we're trying to expand indications
02:32:53beyond lung and bone. We're
02:32:55gonna try to treat multiple
02:32:56metastatic sites at some point
02:32:57eventually. This may be a
02:32:58year or two or three
02:32:59away, from really happening in
02:33:01real time, but we're on
02:33:01the ground level working with
02:33:03the company, working with the
02:33:04with, with the other teams,
02:33:06the UCT Southwestern, City of
02:33:07Hope, Stanford, as well as,
02:33:09the the fifth fifth over
02:33:11at Pittsburgh.
02:33:12So we're working all together
02:33:13on this to, to study
02:33:14outcomes, but also to potentially
02:33:16expand indications and also to
02:33:18study novel isotopes beyond FDG.
02:33:20So, for for that that
02:33:22that really could be very
02:33:22useful down the road.
02:33:24So research opportunities, we have
02:33:25a prospective registry for all
02:33:26of our patients who are
02:33:27getting this right now. We're
02:33:28looking at health related quality
02:33:30of life as well as
02:33:30other endpoints that are the
02:33:32standard endpoints. Also, the punt
02:33:33potential for industry sponsored grants.
02:33:35So we're really looking for
02:33:36for, for collaborations around medical
02:33:38oncology surgery from nuclear medicine,
02:33:40pet center. We're we're very
02:33:42excited for, for this chance
02:33:43to to work with all
02:33:44of you on that for
02:33:45all different disease sites eventually.
02:33:47But then just a quick
02:33:48sneak peek on Proton Center.
02:33:49This is opening in a
02:33:50groundbreaking should be within a
02:33:52month or so, happening in
02:33:53Wallingford. It's a combined effort
02:33:55between a joint venture between
02:33:56Hartford HealthCare
02:33:58and us. So it'll be
02:33:59someone in the middle in
02:34:00Wallingford.
02:34:01And
02:34:02the basic idea here is
02:34:03that, there's a little less
02:34:04load there should be actually
02:34:05substantially less low dose scatter,
02:34:07for for for certain patients
02:34:09who who benefit from that.
02:34:10So, that that could be
02:34:11advantageous to some degree. It's
02:34:13not gonna be for everybody,
02:34:14just like Syntyxt is not
02:34:15for everybody either. So we'll
02:34:30two years. That's the the
02:34:32idea right now.
02:34:34Alright. So thank you very
02:34:35much. Just, really excited to
02:34:36collaborate going forward, and and
02:34:37thank you for everyone's support.
02:34:46Hey, doctor Berg?
02:35:01Sound great. What happened, Henry?
02:35:06No. He's lots
02:35:07go around. It could be
02:35:08under us for all we
02:35:09know.
02:35:11Okay. We're running late, so
02:35:12maybe we'll just skip lung
02:35:13cancer. I don't really care
02:35:14about that.
02:35:16Just kidding. No. So so
02:35:17now we're gonna we're gonna
02:35:18hear about lung cancer. And,
02:35:20no better person to start
02:35:21us off is our,
02:35:23leader of the, of the
02:35:24center, our coleader, was Justin
02:35:26Gosberg and Katie Pilletti, Sarah
02:35:28Goldberg. And Sarah's gonna give
02:35:29us the hot topic. So,
02:35:30Sarah.
02:35:36Thank you so much. Hi,
02:35:37everyone. I think I might
02:35:38have the wrong slides here.
02:35:40I think this is the
02:35:41panel.
02:35:43Here we need Yes. We
02:35:44just need the
02:35:46hopefully, you have my slides.
02:35:49Will not take us out
02:35:50of your time. Oh, thanks.
02:35:51So can I say some
02:35:52things while you're working on
02:35:53that? Yeah. So, it's so
02:35:54wonderful to be here. I
02:35:55was hoping I have an
02:35:56extra minute to say hello
02:35:57and and welcome.
02:35:58It's this is such an
02:35:59amazing event, I think, selfishly
02:36:01mainly because I get to
02:36:02see everyone and so many
02:36:03people that I work with
02:36:04that I don't get to
02:36:05see in person all the
02:36:06time. Do you need me
02:36:06to stop talking to do
02:36:07this? I don't You don't
02:36:08have it.
02:36:10Well, I'll just keep talking.
02:36:12It was not I uploaded
02:36:13them, like, early. That's what
02:36:15happens when you do things
02:36:16early.
02:36:18It doesn't
02:36:20oh, can I give you
02:36:21my laptop?
02:36:22Yeah. You got it. Really?
02:36:24Okay. Hold that thought.
02:36:26Mhmm. Connect your laptop right
02:36:28there. Yes.
02:36:48So so
02:36:49so never to waste an
02:36:50empty microphone.
02:36:52So, you know, lung cancer
02:36:53is one of three areas
02:36:54at Yale where we have
02:36:55a spore, a specialized program
02:36:58of, research excellence,
02:37:00which really takes,
02:37:03science from the labs to
02:37:04clinic. We hope to have
02:37:04this in three or more,
02:37:06but in the next time,
02:37:08or maybe in the next
02:37:09time we meet.
02:37:11Lung cancer,
02:37:12melanoma, and head and neck.
02:37:13You'll hear from all of
02:37:14them today. And,
02:37:15we'll have a couple of
02:37:16very nice cases.
02:37:19And,
02:37:20now to Sarah.
02:37:22I'm going back to my
02:37:23welcome. So anyway, what I
02:37:24was saying before is I
02:37:25really, it's so nice to
02:37:26see everyone. It's, It's, we
02:37:27have our fellows here. I
02:37:28think they're in the, like,
02:37:29the back row. Lots of
02:37:30people from the lung team.
02:37:31It's awesome to see you
02:37:32all here. People from across
02:37:33the network, colleagues, industry partners.
02:37:35So so really wonderful. Now
02:37:37I'm gonna talk about lung
02:37:37cancer. It's not that we
02:37:39have my slides. So I
02:37:40think probably most of you
02:37:41know that there has been
02:37:42a lot happening in the
02:37:44field of lung cancer research
02:37:46in the last several years,
02:37:47but this year was no
02:37:48exception.
02:37:49So I am going to
02:37:50try to summarize that in
02:37:52fifteen minutes as best as
02:37:53I can. Here's what I'm
02:37:54gonna try to cover.
02:37:56I'm gonna focus on early
02:37:57stage and locally advanced non
02:37:58small cell lung cancer. We'll
02:37:59talk about neoadjuvant and perioperative
02:38:01immune therapy. Also talk about
02:38:03how we can incorporate targeted
02:38:04therapies into our management. And
02:38:06then I'm gonna focus on
02:38:07EGFR, ALK, and KRAS in
02:38:08the metastatic non small cell
02:38:09lung cancer space. The initial
02:38:11version of this talk had,
02:38:12like, sixty slides, so that's
02:38:13not possible in fifteen minutes.
02:38:15So this is what I
02:38:15had to omit.
02:38:17There's actually some interesting data
02:38:18this year on oligometastatic disease.
02:38:19I'm not gonna cover that.
02:38:20I'm not gonna talk about
02:38:21small cell. You'll hear a
02:38:22bit about that in the
02:38:23cases. I'm also not gonna
02:38:24cover antibody drug conjugates, a
02:38:26really important area of development
02:38:28in lung cancer, but just
02:38:29ran out of time. So
02:38:30I'll give you a sneak
02:38:31peek at the end about,
02:38:34about another talk I'm gonna
02:38:35be giving that,
02:38:36that you can hear more
02:38:37about those topics. But this
02:38:39is really what I'm gonna
02:38:39try to address today. What
02:38:41is the role for perioperative
02:38:42chemo immune therapy in stage
02:38:43two to three lung cancer,
02:38:45and what about those patients
02:38:46with multi station n two
02:38:47disease? Something that I think
02:38:48has really changed in the
02:38:49last few years.
02:38:50Should we give adjuvant therapy?
02:38:52If we've already given neoadjuvant
02:38:53therapy, how do we use
02:38:54targeted therapy in the early
02:38:55stage locally advanced setting, eGFR,
02:38:58ALK, and KRAS?
02:38:59So first about perioperative
02:39:01immune therapy.
02:39:02So these are the two
02:39:04trials that we have now
02:39:06several years of follow-up data
02:39:07on and are both approved
02:39:09regimens for the treatment of
02:39:11patients with stage two or
02:39:12three non small cell lung
02:39:13cancer. So it's the keynote
02:39:14six seven one and the
02:39:15AGN trial. They're similarly designed
02:39:17studies, although, of course, there
02:39:19are there are differences.
02:39:21The keynote six seven one
02:39:22is, neoadjuvant
02:39:23and then also adjuvant,
02:39:25immune therapy with pembrolizumab.
02:39:27The neoadjuvant portion has chemo
02:39:30with a cisplatin based regimen.
02:39:31This is one of the
02:39:32the Nuance differences in the
02:39:33trials, cisplatin regimens only in
02:39:36in addition to pembro. And
02:39:37with,
02:39:38the Aegean trial, it was
02:39:39any platinum doublet with durvalumab
02:39:42versus placebo and chemotherapy. And
02:39:44so I think we all
02:39:45already know this because now
02:39:46this is a couple years
02:39:46old, this data, is that
02:39:48both of these trials were
02:39:49positive for event free survival,
02:39:51their primary endpoints. And you
02:39:53can see there with several
02:39:54years now of follow-up that
02:39:55the the the benefit persists
02:39:57into three years and beyond.
02:39:59So, again, this was practice
02:40:00changing a couple years ago
02:40:02already and something that's made
02:40:03its way into our management
02:40:04of of patients with lung
02:40:05cancer.
02:40:06We do have some data
02:40:07on overall survival. This is
02:40:08still somewhat, immature analysis, but
02:40:10you can see here that
02:40:11both of these trials look
02:40:12like they are at least
02:40:14trending towards a benefit in
02:40:15terms of overall survival. Both
02:40:17of them are not statistically
02:40:19significant, at least not yet,
02:40:20but we'll we'll keep our
02:40:21eye on that and see
02:40:21if that becomes significant in
02:40:23the future.
02:40:24We also saw data at
02:40:25ASCO on the CheckMate seventy
02:40:27seven t trial. This is,
02:40:28again, a similarly designed perioperative
02:40:30trial, but this time including
02:40:31nivolumab. So patients get four
02:40:33cycles of chemo plus nivolumab
02:40:35and then surgery, and then,
02:40:36nivolumab for thirteen cycles afterwards.
02:40:39This is the the, primary
02:40:40endpoint of event free survival
02:40:42for this trial. And, again,
02:40:43you can see a benefit
02:40:44here. Not as much follow-up
02:40:45here as with those other
02:40:46trials because it's it's more
02:40:47recently,
02:40:48presented, but still a benefit.
02:40:50So now we have three
02:40:51and actually more trials than
02:40:52that that show a significant
02:40:54benefit of adding immune therapy
02:40:55in the neoadjuvant,
02:40:57treatment of, and and add
02:40:59to the perioperative treatment of,
02:41:01stage two and three lung
02:41:02cancer. This is what I
02:41:03really wanted to to drill
02:41:04down on because I think
02:41:05this is something that we
02:41:06started to see this year
02:41:07even more that I think
02:41:08really has changed the way
02:41:09I think about
02:41:10selecting patients for perioperative,
02:41:13therapy
02:41:14is looking at the stage
02:41:15three patients, the n two
02:41:16positive patients. At least in
02:41:17our practice at Yale for
02:41:19years, we really
02:41:21were thinking patients with n
02:41:22two disease, especially multistation n
02:41:24two disease, should be more
02:41:25going the chemoradiation route as
02:41:26opposed to the resection route.
02:41:28But I think this data,
02:41:29as well as some other
02:41:30studies, have really started to
02:41:32show us that these patients
02:41:33can do very well when
02:41:35they get a perioperative approach.
02:41:36So you could see on
02:41:38the bottom left, stage three
02:41:39n two single station, on
02:41:40the right, multi station n
02:41:42two disease. Patients are significantly
02:41:44benefiting from the addition of
02:41:45immune therapy to a perioperative
02:41:47approach. And actually, the hazard
02:41:48ratio for the multi stage
02:41:50and two patients is point
02:41:51two three. So really significant
02:41:53benefit with the addition of
02:41:54immunotherapy for those patients. So
02:41:56to answer my question about
02:41:57early stage non small cell
02:41:58lung cancer, I would say
02:41:59based on the superior event
02:42:00free survival and these trends
02:42:01towards improved OS, I think
02:42:03this is now really our
02:42:04standard of care, perioperative chemoimmunotherapy,
02:42:06either chemopembro, chemo dervo, maybe
02:42:08at some point chemo nivo
02:42:09if that is approved.
02:42:10It really should be considered
02:42:12for patients with stage two
02:42:13to three disease even if
02:42:14there's multi station n two.
02:42:16So what's really critical here,
02:42:17and we try to do
02:42:18this, every week in our
02:42:19tumor board, is to have
02:42:20multidisciplinary
02:42:21discussions. These patients with especially
02:42:23stage three disease, but as
02:42:24also stage two really need
02:42:25to be seeing surgeons, radiation
02:42:27oncologists, medical oncologists upfront and
02:42:29having these kinds of discussions
02:42:31so that we could choose
02:42:32the best therapy.
02:42:34Perioperative
02:42:35chemo immune therapy is
02:42:37clearly an an option for
02:42:38these patients, but is not
02:42:39the only one and and
02:42:40should be at least considered
02:42:41and discussed.
02:42:42So one thing that I
02:42:43think this is, I thought,
02:42:44a really important study, that
02:42:45came out just at World
02:42:46Lung just, a couple weeks
02:42:47ago was trying to answer
02:42:49this question. If we give
02:42:50a patient neoadjuvant therapy, should
02:42:52we then give them adjuvant
02:42:53therapy after surgery? So we
02:42:55don't have prospective data
02:42:58yet. So I'm not sure
02:43:00Maybe one day we will
02:43:01have that. But for now,
02:43:02what we have is this
02:43:03analysis,
02:43:04again, presented at World Lung,
02:43:06which compared the data the
02:43:08patient level data from CheckMate
02:43:09eight one six, which is
02:43:10a neoadjuvant only study, to
02:43:12CheckMate seventy seven t, the
02:43:13trial I just showed you,
02:43:14which was perioperative, and they
02:43:15basically compared how those patients
02:43:16did. So by asking the
02:43:17questions, do you need the
02:43:18adjuvant component?
02:43:19So there are some flaws
02:43:20with this study as as
02:43:21there always will be when
02:43:22you're comparing two trials. Lots
02:43:23of questions that were asked
02:43:25of the speaker,
02:43:26at World Lung. But I
02:43:27think it still starts to
02:43:28get at this question of,
02:43:29is the adjuvant portion important
02:43:31in these patients?
02:43:33And what this study showed,
02:43:34again, with its flaws, showed
02:43:35that it may be actually
02:43:36important. So that what they
02:43:37found was that patients who
02:43:39were on the perioperative regimen
02:43:40with of chemo, NEBO followed
02:43:42by NEBO in the adjuvant
02:43:43setting did better than if
02:43:44they got neoadjuvant alone.
02:43:46I think this is also
02:43:47really important. We often will
02:43:48say, well, what was the
02:43:50pathologic response? Should that guide
02:43:52our use of adjuvant therapy?
02:43:54Again, this is, comparing across
02:43:56trials, so you have to
02:43:57really take that, that as
02:43:59part of the understanding of
02:43:59how to interpret this. But
02:44:01what what they found was
02:44:02that both of these groups
02:44:03appeared to benefit with a
02:44:04hazard ratio point five to
02:44:06point six.
02:44:07Although in and if you
02:44:08look at the curves, it
02:44:09maybe looks like in the
02:44:10non path CR patients, those
02:44:12patients seem to benefit more.
02:44:13These are fairly small numbers.
02:44:14Again, limitations of comparing across
02:44:16trials, but we're starting to
02:44:17see some benefit especially in
02:44:19the non path CR patients.
02:44:21So to answer this question,
02:44:23should we give adjuvant therapy
02:44:25after neoadjuvant therapy? This is
02:44:26something I'm sure we'll debate
02:44:27about for many more years
02:44:28in tumor board, but I
02:44:29would say probably we should.
02:44:30We don't have randomized data
02:44:32to guide us yet, but
02:44:32comparing cross trials and starting
02:44:34to get at it with
02:44:35this this study I showed
02:44:36you, it does seem like
02:44:37maybe there is a signal
02:44:38for benefit of giving the
02:44:39adjuvant component, especially in patients
02:44:42who don't have a past
02:44:43CR.
02:44:44So what about targeted therapy
02:44:45in early stage and locally
02:44:46advanced disease? So we've known
02:44:47this data for some time.
02:44:48This was Adora was the
02:44:50big splash last year at
02:44:51ASCO and several years before
02:44:53that as well. Roy showed
02:44:54us that very clearly in
02:44:55in his present plenary presentations.
02:44:58We know that giving adjuvant
02:44:59osimertinib is beneficial for patients,
02:45:02in after resection. And then
02:45:04we also saw this past
02:45:05year in a publication, New
02:45:06England Journal, that the ALENA
02:45:08trial was also positive. So
02:45:09adjuvant, lectinib, and ALK positive
02:45:11lung cancer patients is also
02:45:13beneficial.
02:45:14What we saw at ASCO
02:45:15this year, and this was
02:45:15in the plenary, got a
02:45:17standing ovation, it was very
02:45:18exciting,
02:45:19was the LAURA trial. So
02:45:20this is
02:45:21giving patients consolidation osimertinib after
02:45:24they receive chemoradiation for stage
02:45:26three unresectable non small cell
02:45:28lung cancer. So these are
02:45:28not the resectable patients. These
02:45:29are patients who are getting
02:45:30chemoradiation.
02:45:31They were randomized to osimertinib
02:45:33versus placebo,
02:45:35and,
02:45:36after they they finished their
02:45:37their chemoradiation. And these are
02:45:39the results. It's probably why
02:45:40I got a standing ovation.
02:45:41Absolutely stunning improvement in progression
02:45:43free survival. A couple of
02:45:44things to note about this
02:45:45study. Patients were given osimertinib
02:45:48indefinitely. It wasn't a designated
02:45:50amount of time. It wasn't
02:45:51one, two, three years. It
02:45:53was forever as as long
02:45:54as it was tolerable and
02:45:55they didn't progress. And so
02:45:57that is really a difference.
02:45:58And and I have to
02:45:59say, when I heard about
02:46:00the design of the study,
02:46:01I was really surprised. But
02:46:02now seeing the results, I
02:46:03think it makes sense because
02:46:05you can see how poorly
02:46:06patients are doing when they
02:46:07don't receive osimertinib. Almost every
02:46:09single patient is progressing in
02:46:11a few years.
02:46:12And so, again, there's limitations
02:46:14in this trial. Like, there
02:46:15there isn't any trial, but
02:46:16really patients with unresectable stage
02:46:18three EGFR immune disease are
02:46:19not doing well,
02:46:21with chemoradiation
02:46:22alone. And so,
02:46:24definite osimertinib does seem to
02:46:26to make sense in these
02:46:27patients after seeing results of
02:46:28this study. I won't go
02:46:29into details, but basically every
02:46:30subset seem to benefit here.
02:46:32And then there's some overall
02:46:33survival data, although very immature
02:46:35and early,
02:46:36with maybe showing a trend
02:46:37towards improvement in overall survival,
02:46:39but we'll probably see more
02:46:40of that in the years
02:46:41to come.
02:46:42So is there a benefit
02:46:44from tardropine
02:46:45early stage and locally advanced
02:46:46disease? Yes. Clearly, for eGFR
02:46:49and ALK.
02:46:50We now use osimertinib. We've
02:46:51been doing this for a
02:46:52few years now for early
02:46:53stage disease. Now we should
02:46:54absolutely be using it after
02:46:55chemoradiation
02:46:56for unresectable disease. Alectinib as
02:46:58well, we're using that now
02:46:59after surgery. We don't have
02:47:01data in the prospect prospective
02:47:02data for using electinib after
02:47:05chemoradiation,
02:47:05but I would argue maybe
02:47:06we should be considering it.
02:47:08What about other subsets like
02:47:10ROS one? Again, we're extrapolating,
02:47:11but I think we should
02:47:12be using it or at
02:47:13least thinking about it for
02:47:14those patients as well where
02:47:15we have really active drugs.
02:47:18And so, really, what this
02:47:18comes down to is we
02:47:20have to test for these
02:47:20alterations, or we'll never know
02:47:22that we have these options
02:47:23for for patients. So we
02:47:24have to test for these
02:47:25molecular alterations, not just in
02:47:27stage four disease, but now
02:47:28in stage two and three.
02:47:29And, actually, the one b's
02:47:30as well is really important.
02:47:31They were included on the
02:47:32trials too in the adjuvant
02:47:34trials. So we need to
02:47:35get that information before we
02:47:37give all of our patients'
02:47:39chemo immune therapy in the
02:47:40adjuvant set. So it's really
02:47:42important to this test
02:47:44before making decisions
02:47:46on
02:47:47moving to advanced disease, and
02:47:49this will be fairly fairly
02:47:50quick, and then we'll wrap
02:47:51it up. So what about
02:47:52for eGFR? A lot is
02:47:53happening in eGFR space in
02:47:54the first line setting. We've
02:47:55seen these two trials over
02:47:56the last maybe two years
02:47:57or so. This is really
02:47:58the now the the contenders
02:47:59for the first line eGFR
02:48:00space, which is a a
02:48:02combination of amavantamab,
02:48:03the eGFR met by specific
02:48:05antibody, plus losertinib, a third
02:48:06generation TKI, and then the
02:48:08FLOR two regimen, which is
02:48:09chemo plus osimertinib. And these
02:48:11two regimens were compared in,
02:48:13obviously, different trials, Mariposa and
02:48:14FLOR two versus osimertinib, and
02:48:16both had significant improvements in
02:48:18progression free survival versus our,
02:48:21to this day at least,
02:48:22standard drug osimertinib.
02:48:23So these are now both
02:48:24approved regimens. You can use
02:48:26these in the first line
02:48:26setting for eGFR patients.
02:48:29Toxicity is absolutely a concern.
02:48:30I'm not putting this here
02:48:31to give you any details,
02:48:32but these are something that
02:48:33needs to be considered when
02:48:34you're thinking about regimens for
02:48:36patients. There's toxicity obviously associated
02:48:38with anything more than a
02:48:39TKI.
02:48:40Emivantamab
02:48:41has unique toxicities of kind
02:48:43of the similar EGFR toxicities,
02:48:45but can't be quite severe
02:48:47in patients. Rash, nail issues,
02:48:50diarrhea. So those are absolutely
02:48:52things to think about. And
02:48:53then chemo, I think we
02:48:54all know well, something to
02:48:55consider for patients. Do they
02:48:56do we wanna bring them
02:48:57back and forth for IV
02:48:58infusions as opposed to an
02:48:59oral drug?
02:49:00So the other thing to
02:49:01think about with these combinations
02:49:02is that there really does
02:49:03seem to be a benefit
02:49:04in the CNS, which is
02:49:05interesting since we don't always
02:49:07think about chemo or an
02:49:08antibody getting into the brain
02:49:10as well. But these drugs
02:49:12really do seem to provide
02:49:13a benefit in terms of
02:49:14time to intracranial progression compared
02:49:16to osimertinib alone.
02:49:18At ASCO, there was an
02:49:19interesting study looking at high
02:49:20risk features. So maybe we
02:49:22can select patients for these
02:49:23combinations based on some other
02:49:25features.
02:49:26And, basically, what they found
02:49:27is all of these high
02:49:27risk features, there seems to
02:49:29be a benefit
02:49:31to giving the combination as
02:49:33opposed to osimertinib alone. So
02:49:34maybe we should use that
02:49:35to select patients. The challenge
02:49:36is almost ninety percent of
02:49:37patients had at least one
02:49:38high risk feature. So you're
02:49:39basically then saying all patients
02:49:41should get combination, and I'm
02:49:42not sure that that's true.
02:49:44We do have some overall
02:49:45survival data. It's not mature
02:49:46yet. I feel like I've
02:49:47said that in multiple studies,
02:49:48but that's how how things
02:49:49are right now. They both
02:49:50seem to be trending in
02:49:51the right direction, but not
02:49:52statistically significant at this point,
02:49:54but maybe one day will
02:49:55be. So how do we
02:49:56select the optimal eGFR strategy?
02:49:58Well, this is from a
02:49:59talk that was given a
02:50:00few or last year at
02:50:01ESMO, but I think it
02:50:02highlights very well the challenges
02:50:03we have. We don't really
02:50:05know. There There's a lot
02:50:06of different options in the
02:50:07first line now and then
02:50:08second and third line where
02:50:09we can be thinking about
02:50:10different strategies for our patients.
02:50:12So how do we select
02:50:13in practice?
02:50:15I think chemo and osmartinib
02:50:16or amylaser really should be
02:50:18considered from really most of
02:50:19our patients if we think
02:50:20that patients can tolerate it,
02:50:22especially if they have high
02:50:23risk features, which again is
02:50:24most of our patients, including
02:50:25brain mets, we should be
02:50:27considering a combination.
02:50:28Single agent OC is still
02:50:30an option,
02:50:31especially if patients are older
02:50:32or frail, low bulk disease,
02:50:33if people really don't want
02:50:35IV infusions or coming back
02:50:36and forth for treatments, absolutely
02:50:38still an option.
02:50:40For ALK, I'll very quickly
02:50:41go through this because I
02:50:42think this is a pretty
02:50:43straightforward,
02:50:45summary, but I think it's
02:50:46really important in case you
02:50:47missed it. This was one
02:50:48of the most exciting studies
02:50:49I thought at ASCO this
02:50:50year. The study the crown
02:50:52study lorlatinib versus crizotinib for
02:50:54first line ALK. Again, an
02:50:55amazing progression free survival curve.
02:50:57This is actually what I've
02:50:58been told. I haven't checked
02:50:59this through, but I've been
02:51:00told this is the best
02:51:01five year
02:51:02PFS for any metastatic lung
02:51:05cancer ever. So sixty percent,
02:51:07of patients are progression free
02:51:09at five years. Really incredible
02:51:11with lorlatinib.
02:51:12Amazing results for intracranial
02:51:14control as well. I won't
02:51:15go into details for the
02:51:16sake of time. Again, something
02:51:18to think about with lorlatinib
02:51:19when you're selecting first line
02:51:20regimens because,
02:51:22alectinib is another very reasonable
02:51:23option for brigatinib as well,
02:51:25is toxicity. So there are
02:51:27adverse events that are pretty
02:51:28unique to lorlatinib, the cognitive
02:51:30side effects, weight gain. So
02:51:31these are some things to
02:51:32consider.
02:51:33But I think one really
02:51:35interesting thing that was shown
02:51:36at ASCO is that even
02:51:38in patients that needed dose
02:51:39reduction, they still did extremely
02:51:41well. And actually, maybe they
02:51:42seem to do better in
02:51:43terms of PFS and time
02:51:44to intracranial progression. So don't
02:51:46be afraid to dose reduce
02:51:47these patients. So very easy
02:51:49answer for me here. What's
02:51:50the best first line treatment?
02:51:51It does seem to be
02:51:52lorlatinib. I've changed my practice
02:51:54based on this data, but
02:51:55you absolutely need to think
02:51:56about side effects discussed with
02:51:57patients. Some patients may not
02:51:58wanna risk the side effects,
02:52:00but I think starting the
02:52:01the drug and really having
02:52:02a low threshold to dose
02:52:03reduce is is a good
02:52:04strategy.
02:52:05In KRAS, we saw that
02:52:07adagrasib does have improved progression
02:52:09free survival compared to docetaxel
02:52:10in the second line setting.
02:52:11We saw similar results with
02:52:13sotorasib a few years ago.
02:52:14These are both now approved
02:52:15drugs and can be used
02:52:16in the clinic.
02:52:18There are some new kids
02:52:18on the block that are
02:52:19coming. We have trials for
02:52:21both of these drugs in
02:52:22the clinic right now in
02:52:23at Yale.
02:52:25Dvoracev and olimerasib, they both
02:52:27look really promising. I'm not
02:52:28gonna go into details, but
02:52:29they look like they have
02:52:30maybe higher response rates and
02:52:31durability than sotorasib or adaggressive
02:52:34and really, really exciting to
02:52:35see them coming into clinic.
02:52:38So I didn't get a
02:52:39chance to talk about so
02:52:40many other really exciting things.
02:52:41All the new agents coming,
02:52:43there were some really interesting
02:52:44and promising drugs that we
02:52:46saw,
02:52:47at ASCO and and and
02:52:48in World Lung and at
02:52:49ESMO. So I had to
02:52:51just I had to get
02:52:51it out there. So we're
02:52:53gonna do this again. I'm
02:52:54gonna do this again. We
02:52:55have a translational lung cancer
02:52:56conference once a month at
02:52:57noon at Yale.
02:52:59And so I'm gonna give
02:53:00you some more updates, the
02:53:01things I didn't get a
02:53:02chance to talk about. We
02:53:03can go into detail and
02:53:04have more of a discussion
02:53:05then. So December fourth at
02:53:06noon. If you're interested and
02:53:08wanna come, let me know.
02:53:08I could send you the
02:53:09link if you don't already
02:53:10have it. Thank you so
02:53:11much.
02:53:20Yeah. Please, George. Okay. That
02:53:22was wonderful. We'll take some
02:53:22questions, but now we're gonna
02:53:24have our panel.
02:53:25So joining the panel, we
02:53:27have doctor Anne Cheng. Where's
02:53:28Anne Cheng?
02:53:31Hi, Anne.
02:53:34And,
02:53:35he does a lot of
02:53:36things, a lot of collaborations
02:53:38and innovative medicine in the
02:53:40swag. Okay. Michael Grant.
02:53:42Michael Grant all the way
02:53:43from Waterbury.
02:53:47Henry Park. We we we
02:53:48met Henry already.
02:53:50Henry, come on over.
02:53:52Andrew Dannesbaum. Is he here?
02:53:54He he did an operation,
02:53:56and he's already back.
02:53:58I hope they're not bleeding
02:53:59or anything.
02:54:01And and then Yu Zheng
02:54:02is not here, so I'm
02:54:03gonna ask Sarah to take
02:54:04her place. Yu Zheng from
02:54:06the VA,
02:54:07if if there's a seat.
02:54:08Okay. So we're we're gonna
02:54:10do a case discussion, and
02:54:11I have to keep us
02:54:12on time.
02:54:15Okay.
02:54:23There's certainly a lot of
02:54:25and that was a slow
02:54:25year for lung cancer.
02:54:29Now where where's Eric?
02:54:31I tell him there's a
02:54:31lot more interesting work in
02:54:32lung cancer than breast cancer.
02:54:35Tell him I said that
02:54:35when he gets back. Okay.
02:54:36So so so now we
02:54:38have our our lung cancer
02:54:39panel, and, we'll go through
02:54:41the case.
02:54:42So we have a sixty
02:54:43five year old woman with
02:54:45a history of, diabetes with
02:54:47sixty,
02:54:48pack years smoking,
02:54:50present with a shortness of
02:54:51breath, chest pain, a cough,
02:54:53a chest x-ray with a
02:54:55large right upper lobe mass.
02:54:56And we can all see
02:54:57that on a PET scan.
02:54:59PET scan with an avid
02:55:00right upper lobe mass, mediastinal
02:55:02lymphadenopathy,
02:55:03right howler, and right supraclavicular
02:55:05lymph nodes.
02:55:06EBAS, with mediastinal lymph node
02:55:08biopsy with, with shows small
02:55:10cell lung cancer. And I'll
02:55:11let you all read that.
02:55:12But, Anne, do you agree
02:55:13that's a good
02:55:15that looks like small cell?
02:55:16You know, pathology says so.
02:55:17It's gotta be true. Pleural
02:55:19fluid cytology is negative for
02:55:20malignancy, and the MRI brain
02:55:22is negative. So, Anne, from
02:55:23what we've seen so far,
02:55:24is this
02:55:26a limited or extensive,
02:55:28situation?
02:55:29Well, this one is a
02:55:30limited stage situation, and it's
02:55:32gonna highlight some of the
02:55:33major advances this year at
02:55:35ASCO. Okay. So it looks
02:55:36pretty limited to me, but
02:55:38okay. Good.
02:55:39So the patient has limited
02:55:41stage I knew that, but
02:55:42I guess small cell lung
02:55:43cancer and at stage three
02:55:45c,
02:55:46t four n three n
02:55:47zero. Radiation oncology can treat
02:55:49with definitive radiation. You can
02:55:50do that, Henry?
02:55:52Gonna use the new machine
02:55:53or just the old stuff?
02:55:56The the the old stuff.
02:55:58The proton won't be ready
02:55:59till twenty twenty six. We
02:56:01gotta wait for that. What
02:56:02initial treatment and radiation schedule
02:56:04would you deliver?
02:56:05One, carboplatin, metoposide, and atezolizumab
02:56:08or divalumab without radiation therapy.
02:56:11Two, cisplatin, metoposide with concurrent
02:56:13definitive radiation followed by observation.
02:56:15And three, cisplatin,
02:56:16metoposide,
02:56:17and concurrent definitive radiation followed
02:56:19by divalumab for two years.
02:56:21So, do we have a
02:56:22way to vote here?
02:56:24Yes.
02:56:25Do you vote? Well,
02:56:28do do people wanna vote?
02:56:30So who who thinks number
02:56:31one? Everyone can vote. And
02:56:32I this is not one
02:56:33of those events where those
02:56:34who aren't health care providers
02:56:35can't vote. Who wants number
02:56:37one?
02:56:38Probably get fired. Who wants
02:56:40number two?
02:56:43Oh, only a few people
02:56:44that can compare me. Who
02:56:44wants number three?
02:56:46Okay. So now we have
02:56:48an August panel. So,
02:56:50doctor Grant, they're they're they're
02:56:51out there in Waterbury. What
02:56:53would you do?
02:56:54Do we have slides? How
02:56:55do you wanna do this?
02:56:56And do we have slides
02:56:57to show?
02:57:00Okay. So so would you
02:57:02like to
02:57:04take a lead?
02:57:06Perfect.
02:57:07So I'm gonna actually turn
02:57:09to Henry to talk about
02:57:10the,
02:57:12the Adriatic
02:57:13trial, and then, I'm gonna
02:57:15talk a little bit about
02:57:16the the sort of implications
02:57:17after that. So Okay. But
02:57:19before we do that,
02:57:20our our public are waiting.
02:57:22So what would you do?
02:57:23Which which option? Three.
02:57:25Michael,
02:57:26Anne,
02:57:27Andrew,
02:57:29Sarah.
02:57:32Is it approved yet?
02:57:34It's it's not approved, but
02:57:36it's it came out in
02:57:37the New England Journal, and
02:57:38we have been using it.
02:57:39I think it's a new
02:57:40standard of care, and we'll
02:57:41we'll look at why. Okay.
02:57:42So everyone's we have unanimous
02:57:44consent on on number three.
02:57:46So okay. Henry, the store
02:57:47is yours, and, let you
02:57:48Hey. I'll just go through
02:57:49these, slides real quick regarding
02:57:51I'll get some immunotherapy trials
02:57:52in limited stage, that have
02:57:54been ongoing, but the the
02:57:55the two that are are
02:57:56have been reported is is
02:57:58it Adriatic,
02:57:59which we'll talk about here.
02:58:00Actually, l u zero zero
02:58:00five is technically embargoed until
02:58:02tonight, so I won't talk
02:58:03about that, right now. But
02:58:05it's gonna be at Astro
02:58:06where I'm I'm going next,
02:58:07would be presented on Monday.
02:58:08Alright. I'll go. I'll turn
02:58:09it off. Okay. So,
02:58:11I mean, so and I
02:58:12could tell you because it's
02:58:13printed in in in,
02:58:27here's where everyone got chemo
02:58:28radiation first, either sixty sixty
02:58:30to sixty six gray,
02:58:32for for,
02:58:33over six weeks or forty
02:58:35five gray BID. So they
02:58:36left a choice, that that
02:58:37was that was open to
02:58:38the investigators.
02:58:39Radiation had to start no
02:58:40later than, say, the end
02:58:41of cycle two of chemotherapy,
02:58:43and then their randomization was
02:58:44either was either immunotherapy or
02:58:46placebo. Now the last part,
02:58:47with this has has hasn't
02:58:48been reported yet, so I
02:58:49won't talk about that yet.
02:58:52Overall survival here. You can
02:58:53see that the the the
02:58:54main differences here, at, the
02:58:56median survival being fifty six
02:58:58fifty five point nine months,
02:58:59for dervva versus thirty three
02:59:01point four, for placebo. So
02:59:02really remarkable differences here in
02:59:05limited stage stage small cell
02:59:05lung cancer where there haven't
02:59:07really been any any advances
02:59:08since the advent of concurrent
02:59:10chemo and radiation together. We've
02:59:12been trying to add dose
02:59:13escalate for radiation. There have
02:59:14been some trials that have
02:59:15looked at that and and
02:59:16so shown overall survival benefit,
02:59:18but not a progression. Free
02:59:19survival benefit, it's it's been
02:59:21kind of messy data. But
02:59:22this is the first real,
02:59:23the
02:59:24the the the the really
02:59:25first real advance, I think,
02:59:26in this area for a
02:59:27very, very long time in
02:59:28this disease.
02:59:29And it really mirrors specific
02:59:31in a lot of ways
02:59:32with non small cell lung
02:59:33cancer in terms of the
02:59:34the kind of ten percent
02:59:35or the absolute difference that
02:59:36you're seeing at both the
02:59:38two year and the three
02:59:39year marks so far. Progression
02:59:40free, you know, same idea
02:59:41here. So it's a pretty
02:59:43remarkable.
02:59:44So, basically, food for thought,
02:59:46you know, when when,
02:59:47when when adding Yeah.
02:59:50I I think that, you
02:59:51know, just to highlight, this
02:59:53is two two years overall
02:59:55survival. Still early, but that
02:59:57kind of a benefit for
02:59:58small cell, I think,
02:59:59everybody can agree, it really
03:00:01impacts their lives. And this
03:00:03was a plenary. There was
03:00:04a standing ovation. And this
03:00:05is something that, you know,
03:00:07is, I think, a real
03:00:09breakthrough in the field.
03:00:10Some of the questions
03:00:12that we are thinking about
03:00:14as as the as we
03:00:15start to implement this, you
03:00:17know, obviously, is anti CTLA
03:00:19four gonna improve and that
03:00:21those that that arm will
03:00:23will read out, eventually.
03:00:26And then to to to
03:00:28emphasize, this is for two
03:00:30years. Right? So this is
03:00:31different from pacific and non
03:00:33small cell, which is for
03:00:34one year. This is dervva
03:00:36for two years.
03:00:37And, you know, should we
03:00:38continue immunotherapy if the patients
03:00:40recur,
03:00:41within two years? So I
03:00:43think that's gonna change our
03:00:45our ideas around platinum chemotherapy
03:00:47challenge and utilization
03:00:49of new agents.
03:00:51We don't know if adding
03:00:52chemo to the IO post
03:00:54progression
03:00:55works. We there is some
03:00:57thought, and and our Insigna
03:00:58trial
03:00:59will read out in terms
03:01:00of saying whether or not,
03:01:01you know, you add chemo
03:01:03and then create more neoantigen
03:01:05antigens. Does that
03:01:08vaccinate does that does that
03:01:10stimulate the immune system?
03:01:13Maybe we can use ctDNA
03:01:14to high identify high risk
03:01:16patients.
03:01:18Small cell is a is
03:01:19a disease which has a
03:01:20high burden of circulating tumor
03:01:22cells, and so maybe we
03:01:23can leverage that. And finally,
03:01:25is limited stage different from
03:01:26extensive stage? Because,
03:01:29you know, in extensive stage,
03:01:31you know that IO only
03:01:32had the benefit adding IO
03:01:33to chemo only had a
03:01:34benefit of about two months,
03:01:36and now we're talking about
03:01:37two years.
03:01:38I think there are other
03:01:40other questions that we can
03:01:41talk about with the panel
03:01:42around PCI and and patients
03:01:44who recur and, the role
03:01:46of surgery for a limited
03:01:47stage small cell.
03:01:51I guess I'll take the
03:01:51first one on PCI. Can
03:01:53you all hear me? With
03:01:54this? Okay. So, with with
03:01:56PCI,
03:01:57you know, that's something that
03:01:58we're we're studying with the
03:01:59MAVERICK trial right now. Thank
03:02:00you for for the slide.
03:02:01So, yeah, if you go
03:02:02back one. So this is
03:02:04a SWOG nineteen it's SWOG
03:02:05eighteen twenty seven, the MAVERICK
03:02:06trial, looking at PCI versus
03:02:08MRI surveillance alone. So on
03:02:10both arms, they get MRI
03:02:11surveillance, but we do have
03:02:12this trial open at Yale
03:02:13still. And and and and,
03:02:14the the target is currently
03:02:15six hundred. We're around two
03:02:17hundred right now overall and
03:02:18but they're trying to to
03:02:19to change the the study
03:02:21design a little bit to
03:02:21get this a little closer
03:02:22to accrual, the target around
03:02:24three sixty instead. But the
03:02:25basic idea is with immunotherapy
03:02:27now, do you need PCI
03:02:29still?
03:02:29And and also with MRI
03:02:31surveillance. And I think, we
03:02:32don't have a slide here
03:02:33on this, but over at
03:02:34ESMO, they presented, the the
03:02:36the actual subgroup analysis of
03:02:38those who did get PCI
03:02:39versus those who did not.
03:02:40And actually those who got
03:02:41PCI and both those who
03:02:42got dervva and those who
03:02:43did not get dervva, the
03:02:44people who got PCI still
03:02:45did better than than those
03:02:47who didn't. Now, of course,
03:02:47that was not a randomized
03:02:48comparison that we really need
03:02:49this trial to to see,
03:02:51but there there was about
03:02:52ten percent difference there in
03:02:53terms of how people did
03:02:54irrespective if if they got
03:02:55immunotherapy. So I still think,
03:02:57I'm not yeah. I I
03:02:58I must still still say
03:02:59standard of care is PCI,
03:03:00but we we are offering
03:03:02this trial to everybody to
03:03:03potentially omit PCI
03:03:05on trial. Let me ask
03:03:06Andrew and get you into
03:03:07the conversation. Is there a
03:03:08role for surgery in limited
03:03:09disease?
03:03:10That's been back and forth
03:03:11in the literature. Yeah. You
03:03:13know, I think this is
03:03:15potentially practice changing from a
03:03:17a surgeon standpoint.
03:03:19Up to this point,
03:03:22typically,
03:03:23surgery has been utilized in
03:03:24the so called very limited
03:03:26stage situation, a small tumor,
03:03:28t one,
03:03:30potentially t two and zero.
03:03:33And,
03:03:34the studies,
03:03:35that had looked at, surgery,
03:03:39was, of the sort of,
03:03:40the decades ago,
03:03:43when there was less,
03:03:45prevalence of the, minimally invasive
03:03:47approach, whether robotic or bats.
03:03:48I think because,
03:03:50a lot of the
03:03:52cancer operation,
03:03:53these days are minimally invasive.
03:03:55Patients are recovering,
03:03:57more quickly,
03:03:58able to get to, back
03:03:59to their, pre, performance status,
03:04:02sooner.
03:04:03I think it's def definitely
03:04:05worth, you know, considering,
03:04:07surgery as a,
03:04:09for a local,
03:04:10control. Right. Does anyone in
03:04:11the audience have questions? Please,
03:04:13I'd love to see interaction
03:04:15rates. I see your fellows,
03:04:16Extra points to fellows asking
03:04:18questions. You can get free
03:04:20lunch next Tuesday.
03:04:22We we already do that.
03:04:24Okay.
03:04:25New stuff. So this patient's,
03:04:26you know, you know, I've
03:04:27been treating this for thirty
03:04:28years, and I I have
03:04:30a few people with small
03:04:31cell limited who are alive,
03:04:32but many will re re
03:04:33re re re re recur.
03:04:34So when they recur, what
03:04:35do we do? After all
03:04:36this, they've already had immunotherapy.
03:04:38They've already had chemotherapy. They've
03:04:39already had radiation. Maybe Adnan
03:04:41just stuck a knife in
03:04:42them. So so then what
03:04:43what's next?
03:04:45Sarah?
03:04:48Feel like I should be
03:04:49the last one to talk
03:04:50about this, but, it should
03:04:51be down at this end
03:04:52of the table. So right.
03:04:53So I think this is
03:04:54really what we wanted to
03:04:55highlight here is that,
03:04:57for really, the only approved
03:04:59drug for years was topotecan.
03:05:02But now we have other
03:05:03options. There's some really exciting
03:05:04drugs that are have just
03:05:06basically shown data. Tarlatanab, I
03:05:07think, is the the main
03:05:08one that we're really excited
03:05:10about. We've had this in
03:05:11trial for the last couple
03:05:12years. We have a few
03:05:13more trials coming with it,
03:05:14but now it's really standard
03:05:16of care for our patients,
03:05:17select patients. I think I'll
03:05:18let Anne tell you more
03:05:19about how to select these
03:05:20patients, and she has a
03:05:21lot more experience than I
03:05:22do with this.
03:05:23But it's so exciting that
03:05:24we have this option for
03:05:25our patients. And then the
03:05:26antibody drug conjugates are also
03:05:28looking pretty good in small
03:05:29cell. I I I was
03:05:30really surprised to see that.
03:05:32We again, we've had a
03:05:33couple of trials with them.
03:05:34We have a few more
03:05:34coming, and we're seeing great
03:05:36responses durable with ADC. So
03:05:38those are not approved in
03:05:39small cell yet, but hopefully
03:05:40that will be to come.
03:05:42Right. You know, the clinical
03:05:43trials of today are the
03:05:44therapies of tomorrow and when
03:05:45terlotinib had its approval. And
03:05:47Kuniculata had one of our
03:05:48patients, you know, front and
03:05:49center in the New York
03:05:50Times. Small little paper there.
03:05:51Right? Yeah. We we prefer
03:05:53the New Haven register where
03:05:53we got the New York
03:05:54Times.
03:05:55Playing playing pickleball. I mean,
03:05:57this is what's so exciting
03:05:58about terlotinib, this bispecific, is
03:06:00that it directs,
03:06:02it it leverages what we
03:06:04know about small cell biology.
03:06:05These are cold tumors. There's
03:06:07not a lot of t
03:06:08cells there. So having a
03:06:09bispecific that one head is
03:06:11recognizing d l l three
03:06:13on the tumor cell and
03:06:14the other recognizing c d
03:06:16three to bring your immune
03:06:17system in, that it's so
03:06:19exciting to see response rates
03:06:20of forty percent stable disease,
03:06:22thirty percent, so seventy percent
03:06:24disease control rate. And at
03:06:26six months, sixty percent of
03:06:28the patients are still on
03:06:30on on the on the
03:06:31drugs. So we're still the
03:06:33median is somewhere probably gonna
03:06:34be somewhere around eleven eleven
03:06:36months,
03:06:37which which is terrific for
03:06:39for this patient population.
03:06:41There there's really high cytokine
03:06:42release syndrome. So for the
03:06:44first two doses,
03:06:45day one, day eight, we
03:06:47do,
03:06:49admit them to a hospital
03:06:50after afterwards for for monitoring.
03:06:52But at ESMO IO, we
03:06:54hopefully are gonna be presenting
03:06:56our data,
03:06:57at,
03:06:59around the eight hour monitoring.
03:07:00So so we stay tuned
03:07:01for that.
03:07:03Hopefully, it'll get accepted and
03:07:04I can go. But Okay.
03:07:05Okay. We have a lot
03:07:06of trials, and and we're
03:07:07really excited about this space.
03:07:09So if you have patients,
03:07:10please please send them or
03:07:12or give us a call,
03:07:14email, and and we'd be
03:07:15happy to help. K. Good.
03:07:16We only have a few
03:07:17minutes, but we'll try to
03:07:18quickly do the second case.
03:07:19Okay? And we might just
03:07:21limit who talks just so
03:07:22we can move it along
03:07:23because otherwise, we'll miss lunch.
03:07:24You know?
03:07:25Okay. So so, Mike, do
03:07:26you wanna tell us about
03:07:27the case? Yes.
03:07:29You hear me?
03:07:31So this is case number
03:07:32two.
03:07:33It's a seventy five year
03:07:34old male
03:07:35with a history of lymphoma,
03:07:38and also a smoking history,
03:07:39but, you know, remote, quit
03:07:41in the nineteen nineties.
03:07:44This patient recently
03:07:45developed shortness of breath and
03:07:47a cough
03:07:48and had a chest a
03:07:49chest CT, a chest X-ray
03:07:51first, but then a chest
03:07:52CT, which showed a right
03:07:53upper lobe mass of, three
03:07:56point five centimeters
03:07:57and an enlarged paratracheal lymph
03:07:59node.
03:08:00You click, Roy, you'll see
03:08:02the PET scan. So the
03:08:03the PET scan showed that
03:08:05this right upper lobe mass
03:08:06was hypermetabolic, and there was
03:08:07a paratracheal lymph node also
03:08:09hypermetabolic,
03:08:11and then a subcarinal
03:08:12lymph node,
03:08:14and a and a hilar
03:08:15lymph node as well, but
03:08:16nothing outside of the chest.
03:08:17So this looks to be
03:08:19locally advanced disease.
03:08:22So the the rest of
03:08:23the workup ensued, good pulmonary
03:08:25function.
03:08:28He ended up undergoing EBUS,
03:08:30which showed lung adenocarcinoma
03:08:33in the paratracheal and hilar
03:08:34lymph nodes.
03:08:36But at that time, the,
03:08:38the biopsy was insufficient for,
03:08:40tissue testing or molecular testing.
03:08:42MRI brain was negative, and,
03:08:45it was decided to undergo
03:08:46a repeat biopsy to get
03:08:48enough tissue to undergo the
03:08:49molecular testing that we all
03:08:51want.
03:08:52And
03:08:54then, this molecular testing showed
03:08:56a PDL one of fifty
03:08:57percent and an EGFR exon
03:08:58nineteen deletion.
03:09:00And this is a an
03:09:01exon nineteen deletion that is
03:09:03uncommon, but
03:09:04is known to be sensitive
03:09:06to TKIs, to all TKIs,
03:09:07but,
03:09:08most notably in this case,
03:09:10osimertinib.
03:09:11So this was a stage
03:09:12three a,
03:09:14as Sarah was talking about,
03:09:15multi station
03:09:16n two disease.
03:09:18You know, this is obviously
03:09:19a patient that we
03:09:21need to talk about in
03:09:22the context of a multidisciplinary
03:09:24discussion with radiation, medical oncology,
03:09:27surgery, you know, all the
03:09:28disciplines that go into lung
03:09:30cancer treatment. We got a
03:09:31whole tumor board here. Okay.
03:09:33So Yeah. Andrew, is is
03:09:34this a surgically resectable case?
03:09:38You know, I I think,
03:09:41for for this situation
03:09:43where,
03:09:44I guess, we'll maybe talk
03:09:45in the next slide, as
03:09:47far as the, therapies for
03:09:48multidisciplinary
03:09:49care. You know, I think,
03:09:52it's,
03:09:53number one, resectable from a
03:09:55tumor standpoint.
03:09:57Number two, you know, evaluating
03:09:59the,
03:10:00peritracheal
03:10:01nodes.
03:10:02I think that'll be important
03:10:04as far as,
03:10:06evaluating whether it is, bulky,
03:10:09too bulky to,
03:10:12to be able to get
03:10:13a negative margin, or not.
03:10:15So I I think there
03:10:16is some,
03:10:18it's sort of a case
03:10:19by case here. So, Henry,
03:10:20I know you already ate
03:10:21anything, but but what what
03:10:23do you think about this
03:10:24one? Well, here, I think
03:10:25that well, here here the
03:10:27patient, you know, it'll really
03:10:29depend on the surgeon. And
03:10:30and that is I think
03:10:32the key is that, if
03:10:33if the patient is is
03:10:35it's medically operable and the
03:10:36the surgeon believes they could
03:10:37do the resection upfront, then
03:10:38I believe they they they
03:10:39should get chemo IO first
03:10:41and then surgery. I think
03:10:42the the question is when
03:10:43it's more borderline,
03:10:44and the pay and the
03:10:45surgeon is not sure if
03:10:46the if the patient will
03:10:47be able to to get
03:10:48surgery, you know, after the
03:10:49the chemo immunotherapy, especially if
03:10:51it stays stable. But this
03:10:52guy has
03:10:53has a, EGFR nineteen. So
03:10:55it's probably gonna either be
03:10:56surgery, phybe, action therapy, or,
03:10:59I'm sorry. I just have
03:10:59to do CGM. Or chemo
03:11:00radiation, which I I actually
03:11:01I'm kidding you. I actually
03:11:02am a big fan of
03:11:03chemo radiation. I trained for
03:11:04Farber. We used to do
03:11:05a lot of that. For
03:11:06the EGR patients, I mean,
03:11:07we we we see the
03:11:07outcomes from Laura now that
03:11:09that show that with chemoradiation
03:11:10after osmirnib,
03:11:11you know, the the results
03:11:12are overall pretty reasonable, but
03:11:13I I would say here
03:11:14to go for surgery first.
03:11:15This could be either an
03:11:16Adora or a Laura. Exactly.
03:11:18So what do you name
03:11:19your daughter? So so so
03:11:21who wants to take Anne?
03:11:21What will you do here?
03:11:23Give it to
03:11:25alright. I I I Mike
03:11:26or That's a good one.
03:11:27I just thought of that.
03:11:28So the,
03:11:29the, I think this this
03:11:31case is highlights a couple
03:11:33of different things that are
03:11:34important. Number one,
03:11:36you know, that we agree
03:11:37that there's a
03:11:38a potentially
03:11:39cure of curative intent,
03:11:42therapy available for this patient.
03:11:44So the best way to
03:11:45make that possible for the
03:11:46patient is obviously the way
03:11:47that everybody wants to go.
03:11:51The other thing is going
03:11:52back and biopsying to get
03:11:53this biomarker. You know, Sarah
03:11:55mentioned the importance of knowing
03:11:57upfront so that we can
03:11:58select the best patient the
03:12:00best treatment for the patient.
03:12:02And then third, you know,
03:12:04the PDL one fifty percent
03:12:06is entice or, you know,
03:12:08tempting to look at, but
03:12:09in the context of the
03:12:11other genomic alterations is really,
03:12:14what's really important for selecting
03:12:15the best treatment. So the
03:12:17EGFR mutation,
03:12:19really, I think, takes the
03:12:20immunotherapy
03:12:21options,
03:12:22makes them less,
03:12:24less likely to to be
03:12:26the way we go because
03:12:27there's targeted therapy options that
03:12:29have resulted recently. So,
03:12:33and then yeah. So the
03:12:34the the I think the
03:12:35best role for this patient,
03:12:37if if surgery is not
03:12:38the best way to go
03:12:39up front, then
03:12:41I think chemoradiation
03:12:42seems like a a reasonable
03:12:44option. And, obviously, tigrisso or
03:12:46osimertinib would be used afterwards.
03:12:48I like this. I can
03:12:50okay. So, show of hands
03:12:52for my panel. Who's gonna
03:12:53go Laura?
03:12:57Who's gonna do the Adora?
03:12:59This is why you have
03:13:00two or one. And, of
03:13:00course, you'll ask the patient
03:13:01what he or she thinks.
03:13:02Right?
03:13:03Any comments? This is a
03:13:04this was a good case.
03:13:05Good.
03:13:08Cheers. Anyone have any any
03:13:10thoughts?
03:13:11So it just it just
03:13:12shows that in this day
03:13:14and age, there are a
03:13:15lot of choices. And,
03:13:17you wouldn't even know about
03:13:18this choice, right, if we
03:13:19didn't profile the patient. So,
03:13:21Sarah, is everyone profiled before
03:13:22they get to tumor board?
03:13:25Yeah. I think sometimes we
03:13:26don't have enough tissue. I
03:13:27think sometimes it's we may
03:13:29not have even it's supposed
03:13:30to be reflex. It's not
03:13:31always. Right? Sometimes things happen.
03:13:33But I think this is
03:13:34an amazing case, Mike, that
03:13:35you brought this up that
03:13:36you redid it up you
03:13:36did another biopsy because of
03:13:38how important that is. So
03:13:39I it we they should
03:13:40be. And if they're not,
03:13:41it's it's really important to
03:13:42make that decision. And I'll
03:13:43tell you, I've worked at
03:13:44three, maybe four centers in
03:13:46my career. It would be
03:13:47different in each one. You
03:13:47know, that depends who your
03:13:48radiation oncologist is and your
03:13:50surgeon and but but really
03:13:51and things change. But, you
03:13:53know, the the best therapy
03:13:54is, I I believe, when
03:13:55you decide what you're gonna
03:13:56do upfront. The problem is
03:13:57if you then start having
03:13:58to do split course and
03:14:00that's not good. Right?
03:14:01But, we're out of time,
03:14:02but you can ask our
03:14:03experts later, you know, their
03:14:05their thoughts. Actually, the photographer's
03:14:07here, so let's let's get
03:14:08a nice picture of this
03:14:08this this panel.
03:14:11But, thank you all very
03:14:12much. Stay for your photo,
03:14:13and and then we'll we're
03:14:14gonna move on to the
03:14:15rapid fire here. K. Thank
03:14:17you.
03:14:22Don't forget the moderator too
03:14:23over here. K. Okay. Okay.
03:14:24Now now
03:14:26now now now in the
03:14:27in the
03:14:28in
03:14:29the neck I need it
03:14:31for my CME report because
03:14:32they tell me I don't
03:14:33have my CME or my
03:14:34mock. We're actually getting mock
03:14:35points for anyone who needs
03:14:37mock. You know, that that's
03:14:39a real pain in the
03:14:40neck. Okay. So now we
03:14:42have five rapid fire report,
03:14:43rapid shots. Practice changing literature
03:14:45across disciplines. And I'm gonna
03:14:47actually be strict here because,
03:14:49if we have fifty minutes
03:14:51now plus maybe a couple
03:14:52of minutes for the changeover,
03:14:53that'll get us to lunch
03:14:54about twelve twenty,
03:14:56twelve twenty five. And look
03:14:57how beautiful it is out
03:14:58there. It'll be very nice
03:15:00to have some lunch. So
03:15:01I'm gonna keep the clock
03:15:02and,
03:15:03Harriet, you're the first up
03:15:04doctor Harriet Kluger.
03:15:06She's a vice chair of
03:15:07medicine. She's an associate director
03:15:09in the cancer center.
03:15:10She's the leader of the
03:15:11melanoma program and the the
03:15:12melanoma spore. She's gonna give
03:15:14us ten minutes on melanoma
03:15:16rapid shot, Harriet. And and
03:15:18Nelson's gonna help you out
03:15:19here. Hey. Thank you.
03:15:23Thank you. It's it's Nelson.
03:15:27Good morning, everyone, and I
03:15:29will try to be as
03:15:30rapid as possible. So
03:15:32ten minutes on melanoma. What's
03:15:34changed? So
03:15:35the bottom line is metastatic
03:15:37melanoma,
03:15:38not much changed in a
03:15:39big way this year. We've
03:15:41had many, many updates in
03:15:42recent years, but what I
03:15:43will talk to you about
03:15:44is adjuvant and neoadjuvant
03:15:46therapy.
03:15:48So,
03:15:50I'm gonna start off with
03:15:52one of the late breaking
03:15:53abstracts from ASCO
03:15:55where, where we presented a
03:15:56study of neoadjuvant followed by
03:15:58adjuvant therapy in high risk
03:16:00stage three b melanoma
03:16:02where the results were quite
03:16:03staggering,
03:16:04but one needs to interpret
03:16:06them with some caution. This
03:16:07was presented by my colleague,
03:16:09Christian Blanc. And the second
03:16:10one, I actually had to
03:16:12select from a plethora of
03:16:13novel therapies that are coming
03:16:15through the pipeline,
03:16:16and I picked one, a
03:16:18new drug called preventofusp,
03:16:20which I think might actually
03:16:22make it all the way
03:16:23through the pipeline, and that'll
03:16:24be for metastatic melanoma.
03:16:26So the NEDENA study was
03:16:28a very large multi multi
03:16:29site investigator initiated trial, essentially
03:16:32the type of study that
03:16:33all of us in academia
03:16:34dream of doing. It was
03:16:36a study for macroscopic stage
03:16:37three melanoma, which was defined
03:16:39as palpable melanoma.
03:16:41Because it was an investigated
03:16:43initiated study, provides a terrific
03:16:45opportunity
03:16:46to interrogate samples in death
03:16:48in-depth.
03:16:49And it accrued mostly in
03:16:50Europe and Australia in a
03:16:52mere seventeen months, which is
03:16:53fairly unheard of for a
03:16:55disease that's not particularly common.
03:16:57The rationale behind doing this
03:16:59trial was that there was
03:17:00another study that was led
03:17:01by Swag that I discussed
03:17:03here. I can't remember if
03:17:04it's last year or the
03:17:05year before, and I'm sure
03:17:06everyone remembers exactly what I
03:17:08said, of course. But we
03:17:10discussed the study of pembro
03:17:11that was given before surgery
03:17:13for about three cycles followed
03:17:15by surgery and then adjuvant
03:17:16pembro for up to a
03:17:17year, and that appeared to
03:17:18be superior to adjuvant pembro
03:17:21alone.
03:17:22So the group in Europe
03:17:23took a look at the
03:17:24armamentarium of drugs that we
03:17:25have, and they've they did
03:17:26a few very small,
03:17:28pilot studies looking at neoadjuvant
03:17:30ipinivo,
03:17:31which as you all know
03:17:33is much more active
03:17:34than anti PD one alone.
03:17:37The, early studies provided the
03:17:39biological rationale and some safety
03:17:41data. And and most importantly,
03:17:43they provided the support for
03:17:44pathology driven postoperative
03:17:47therapy.
03:17:47So the study design is
03:17:49over here on the right
03:17:50in a fairly simplified fashion.
03:17:52The experimental arm got a
03:17:54flat dose of epilimumab
03:17:55approximately equivalent to one milligram
03:17:57a kilogram along with nivolumab
03:18:00more or less three milligrams
03:18:01a kilogram
03:18:02for two cycles.
03:18:04Patients then on then went
03:18:05on to have a nodal
03:18:06dissection.
03:18:07If they had a major
03:18:09pathologic response, which was very
03:18:10carefully designed by the pathologists
03:18:12who worked in the study,
03:18:14they then just went into
03:18:15follow-up and didn't get any
03:18:17additional therapies. So in essence,
03:18:18the intent was to cure
03:18:20with two cycles of treatment
03:18:22only.
03:18:23If they did not have
03:18:24a major pathological
03:18:25response, they then went on
03:18:27to get a BRAF MAC
03:18:28inhibitor if the tumors harbored
03:18:30a BRAF mutation or eleven
03:18:31months of nivolumab.
03:18:34The comparator arm was a
03:18:35simple twelve month of nivolumab
03:18:38after nodal dissection.
03:18:40And the key result is
03:18:41depicted in the slide over
03:18:42here. There was a staggering
03:18:43difference
03:18:44in the primary endpoint, which
03:18:46was event free survival
03:18:48at twelve months. Now remember,
03:18:49twelve months is quite early.
03:18:51But the neoadjuvant group of
03:18:52patients had an eighty three
03:18:54percent eighty three percent
03:18:56incidence of event free survival,
03:18:59compared to the adjuvant arm
03:19:01with fifty seven percent.
03:19:04So why do we actually
03:19:05give systemic therapy in resectable
03:19:07stage three melanoma? Because in
03:19:09theory, you just take out
03:19:09the tumor and you're done.
03:19:11Well, obviously, you want to
03:19:13increase the cure rate by
03:19:14eliminating microscopic disease that's not
03:19:16seen on scans at the
03:19:17time. But we gotta be
03:19:19careful to do this while
03:19:20minimizing toxicity, be it physical,
03:19:22emotional, or financial. And we
03:19:24also wanna decrease the recurrence
03:19:26rate. But overall, we really
03:19:27wanna increase the cure rate,
03:19:29and that's why we have
03:19:30to keep our eye on
03:19:31the prize in this era.
03:19:33There's a list of potential
03:19:34drugs that have been studied
03:19:35in the adjuvant setting. I
03:19:36have them in that middle
03:19:38bullet over here. But we
03:19:39take we choose the drugs
03:19:41based on what we know
03:19:42from the macrometastatic
03:19:43setting, where we know that
03:19:45dual checkpoint inhibitors
03:19:47confer better response rates than
03:19:49monotherapy,
03:19:50and that immunotherapy might actually
03:19:52cure patients, whereas targeted therapies
03:19:55don't. But we've also got
03:19:56to take into account that
03:19:57the immune related adverse events
03:19:59are a challenge.
03:20:00So drugs that have been
03:20:02used over the years, interferon,
03:20:04no longer used. Eplimumab,
03:20:06no longer used because anti
03:20:07PD one is better and
03:20:08less toxic.
03:20:10There was a study of
03:20:11epi and nivo in the
03:20:12adjuvant setting. It was not
03:20:14superior to nivo alone. And
03:20:16BRAF and MEC inhibitors are
03:20:18still used because they don't
03:20:19have long term side effects
03:20:20or long term sequela,
03:20:22but there's no actual improvement
03:20:25in survival in any of
03:20:26them.
03:20:27So even if you look
03:20:28at anti PD one alone,
03:20:29if you give adjuvant pembrolizumab
03:20:32compared with standard of care,
03:20:33which at the time was
03:20:34observation,
03:20:35you can see that the
03:20:36survival curves are completely overlapping.
03:20:39So adjuvant therapy
03:20:40is no better than rescuing
03:20:42a patient if you just
03:20:43do the surgery, let things
03:20:45grow, and then treat them
03:20:46at the time of recurrence
03:20:47in terms of overall survival.
03:20:49So it's not clear that
03:20:51we need to actually intervene
03:20:53early at all.
03:20:55So
03:20:55the the DINIS trial wasn't
03:20:57actually intended to address this
03:20:59particular question of whether we
03:21:01need to intervene early. It
03:21:02was intended to address the
03:21:03question of what early intervention
03:21:05is superior.
03:21:07So just to summarize the
03:21:08the prior trial of neoadjuvant
03:21:10follow followed by adjuvant anti
03:21:12PD one alone, we do
03:21:13see a difference
03:21:15in progression free survival. It's
03:21:16not all that clear that
03:21:17overall survival is impacted at
03:21:19all, and that's the curve
03:21:20on the right.
03:21:21And the reason that we're
03:21:22so hesitant,
03:21:24to give immunotherapy when we
03:21:25don't need to is the
03:21:26immune related adverse events. So
03:21:28on this particular trial, the
03:21:30NADINA study,
03:21:31the, there was one death
03:21:33actually on the NEBA only
03:21:34arm from pneumonitis.
03:21:36The incidence of grade three
03:21:38immune related adverse events was
03:21:39doubled. But that in itself
03:21:40isn't all that important if
03:21:42we can rescue a patient
03:21:42with steroids for a few
03:21:43weeks.
03:21:45But we need to really
03:21:46think about those that are
03:21:47irreversible. So adrenal insufficiency,
03:21:49seven percent on the combination
03:21:51therapy arm, one percent on
03:21:52the adjuvant anti PD one
03:21:53only arm. If you take
03:21:55a patient who's young or
03:21:56not young, it doesn't really
03:21:57matter, but person who's going
03:21:59to who's expected to live
03:22:00for many years and you
03:22:01you make the render them
03:22:03adrenal insufficient for the rest
03:22:04of their life, that actually
03:22:05really impacts quality of life.
03:22:07There are others that weren't
03:22:09even documented over here, type
03:22:10one diabetes, cardiotoxicity,
03:22:12and neurologic toxicities.
03:22:14Those oftentimes are not reversible.
03:22:16So the main findings of
03:22:18the NEDENA study, just in
03:22:19summary, are that there was
03:22:21this highly statistically significant benefit
03:22:23in event free survival in
03:22:25the combination epinevo,
03:22:27pre op compared to standard
03:22:30standard adjuvant post op,
03:22:32but
03:22:33the data are still early.
03:22:35And it's not clear that
03:22:36this confirms an overall survival
03:22:39benefit in the long run.
03:22:41It this this approach is
03:22:43good in that it's very
03:22:44brief. You treat patients for
03:22:46three cycles, and you could
03:22:46then stop if they have
03:22:48a good response. And it
03:22:49and it works for patients
03:22:50or it it it was
03:22:51beneficial for patients whether or
03:22:53not they had a BRAF
03:22:55mutant tumor. But, again, remember
03:22:57that the incidence of immune
03:22:58related adverse events needs to
03:23:00be be weighed very carefully
03:23:01against the benefits,
03:23:03particularly as we still don't
03:23:04have clear evidence that there's
03:23:06an overall survival benefit.
03:23:08So to move on to
03:23:09the metastatic setting,
03:23:11ipilimumab and nivolumab have really
03:23:12changed what we do. At
03:23:14least fifty, sixty percent of
03:23:15our patients are alive after
03:23:17five years,
03:23:19And our goal is to
03:23:20keep improving that survival benefit.
03:23:22So we've heard over the
03:23:23last few years about,
03:23:25the I events adoptive cell
03:23:26therapy, which might cure a
03:23:28few a few more patients,
03:23:29like another ten, fifteen percent
03:23:30of patients
03:23:32who, aren't cured from ipinivo,
03:23:34but we still need a
03:23:35whole list of things that
03:23:36we can do for patients
03:23:38who don't respond to those
03:23:39initial therapies. So I selected
03:23:41one therapy only among the
03:23:43list of things that were
03:23:44presented at, at ASCO
03:23:47just because my gut feeling
03:23:48is that this one might
03:23:49actually make it all the
03:23:50way through.
03:23:51So this was a phase
03:23:52one study
03:23:53of a, CD three,
03:23:56country it's a CD three
03:23:57combo therapy that targets PRAME,
03:23:59a bispecific
03:24:00antibody in the post checkpoint
03:24:02setting for cutaneous melanoma.
03:24:05So brinetofas
03:24:07is made by EMCOR.
03:24:08It's,
03:24:10it's essentially
03:24:11a t cell receptor that
03:24:13recognizes PRAME when when PRAME
03:24:16is,
03:24:16is presented within the context
03:24:18of HLA a two zero
03:24:19one. Now this can be
03:24:20made for all sorts of
03:24:21HLA sub, subsets, but a
03:24:23two zero one is the
03:24:24most common
03:24:26HLA subtype that's, that's seen
03:24:28in Caucasian patients who are
03:24:30those that tend to have
03:24:31melanoma.
03:24:33So when PRAME is presented,
03:24:36by the t
03:24:37by the HLA molecules, it's
03:24:38recognized by this by,
03:24:41by brinetofas.
03:24:42And then brinetofas, on the
03:24:44other side, has an antibody
03:24:45to CD three. So, essentially,
03:24:46the CD three cells are
03:24:47brought into the tumor microenvironment.
03:24:49We just heard something similar,
03:24:51in
03:24:52in small cell lung cancer.
03:24:54So this is this is
03:24:55a new class of drugs,
03:24:57and I think we're gonna
03:24:58be seeing more and more
03:24:59of these in various different
03:25:00clinical settings.
03:25:03So for melanoma specifically,
03:25:06there was a monotherapy
03:25:07arm and a and a
03:25:08combination with pembro. I'm not
03:25:09gonna talk about the combination
03:25:10with pembro because the and
03:25:11the patients was too small.
03:25:12But if you just follow
03:25:13the monotherapy arm over here,
03:25:15forty seven patients in expansion
03:25:16cohorts
03:25:17were treated with brinetofas
03:25:19per two dose levels.
03:25:20And taken in combination,
03:25:22the overall response rate was
03:25:24only thirteen, fourteen percent. But
03:25:25if you combine the part
03:25:27the partial response with stable
03:25:29disease with confirmed tumor reduction,
03:25:31which in melanoma is somewhat
03:25:33meaningful,
03:25:33you do see that we're
03:25:34getting around a twenty five
03:25:36to thirty percent actual benefit
03:25:38in patients who are PRAME
03:25:39positive.
03:25:40None of the patients whose
03:25:41tumors did not express PRAME
03:25:43responded. In other words, you
03:25:44need PRAME on the tumor
03:25:46surface in order to respond.
03:25:49The clinical benefit can be
03:25:51prolonged. As you see in
03:25:51the spider plot over here,
03:25:53the vast majority of the
03:25:54responses were ongoing. I think
03:25:56I'm being told to want
03:25:57to
03:25:58wrap it up. Okay. So
03:26:00this was very well treated.
03:26:02It's early in the game.
03:26:03We do believe that this
03:26:04is this is now going
03:26:05into phase three and more
03:26:07to come in the coming
03:26:08years. And with that, I'll
03:26:09stop and let the next
03:26:10rapid shot
03:26:12hot shot.
03:26:21Okay. It's a little too
03:26:22early for those kind of
03:26:23shots. Don't drop it. Okay.
03:26:24So so next rapid shot.
03:26:26Hi, Nick. We we we
03:26:27have to keep on time.
03:26:28So Nicholas Blondin,
03:26:29wonderful neuro oncologist, is gonna
03:26:31tell us about neuro oncology.
03:26:32Nick.
03:26:45Thanks. Thanks to doctor Herbst
03:26:46for the invitation to present
03:26:48some abstracts from ASCO,
03:26:50in the neuro oncology section.
03:26:52Nick Blondin, neuro oncology.
03:26:55No conflicts of interest. I
03:26:57was an investigator for the
03:26:58three studies I'm gonna be
03:26:59presenting on enrolled patients into
03:27:00all these studies. I'll be
03:27:02talking about,
03:27:03study for GBM, meningiomas,
03:27:05and brain metastasis.
03:27:07The first study that I'm
03:27:09presenting is, the evaluation of
03:27:10val o eight three in
03:27:11the GBM AGILE study, which
03:27:13is a phase three registration
03:27:15platform trial for newly diagnosed
03:27:16and recurrent glioblastoma.
03:27:19The goal of the GBM
03:27:20AGILE
03:27:22platform is, an adaptive design
03:27:24to test multiple therapies against
03:27:26one standard control arm for
03:27:28newly diagnosed and recurrent glioblastoma
03:27:30patients. The control arm is
03:27:32radiation with temozolomide,
03:27:34and the, recurrent,
03:27:37is with lamustine.
03:27:39Both temozolomide and lamustine are
03:27:40the standard of care, but
03:27:41relatively ineffective chemotherapies for glioblastoma.
03:27:44This reports the VALO eight
03:27:45three arm.
03:27:47Other arms are currently enrolling,
03:27:48including VT ten twenty one,
03:27:50travailazole, and ADIPEG twenty.
03:27:53This drug, the, VALO eight
03:27:54three has been,
03:27:56in development for some years.
03:27:57It's a non selective alkylating
03:27:59drug.
03:28:00It acts independently of MGMT
03:28:02methylation status. It was studied
03:28:04in, some small studies suggesting
03:28:06efficacy in GBM.
03:28:07So I was studied, in
03:28:09this study, but unfortunately, there
03:28:10was no improvement
03:28:12in overall survival in, any
03:28:15GBM population,
03:28:16newly diagnosed,
03:28:18patients that were methylated or
03:28:19unmethylated
03:28:20or recurrent patients.
03:28:22Unfortunately,
03:28:24there was
03:28:25these are the standard,
03:28:27you know, survival curves of
03:28:28GBM,
03:28:29a disease where survival is
03:28:30measured, in months,
03:28:32sometimes even in weeks.
03:28:35So the,
03:28:37key takeaway was that the
03:28:39platform is working well. I
03:28:40mean, from start to finish,
03:28:41this was less than three
03:28:42years to get this data,
03:28:44which is much faster than
03:28:46other studies in glioblastoma
03:28:48historically.
03:28:49The drug didn't improve
03:28:50survival compared to, control,
03:28:54but this study continues ongoing.
03:28:56Hopefully, a future arm will
03:28:57be effective.
03:28:59Alright. In the, second,
03:29:01study, this was, an aligned
03:29:02study, a phase two trial
03:29:03of abemaciclib
03:29:05in patients with refractory meningiomas
03:29:07harboring mutations in NF two
03:29:09or CDK pathway alterations.
03:29:12It was found that,
03:29:14meningiomas
03:29:15have several driver mutations.
03:29:17A small majority,
03:29:18like around fifty to sixty
03:29:20percent of meningiomas harbor loss
03:29:22of CDKN two a,
03:29:24and NF two.
03:29:25So using drug like a
03:29:27targeted therapy for this seemed
03:29:28to make sense.
03:29:30CDK inhibitors are used in
03:29:31other cancers.
03:29:33So this, aligned study is,
03:29:35a platform study for patients
03:29:37with recurrent or progressive meningiomas,
03:29:40with different molecular subtypes being
03:29:42treated with different drugs.
03:29:44This again was for patients
03:29:45with NF two or CDK
03:29:47alteration.
03:29:49And,
03:29:50it was a single agent
03:29:51study.
03:29:52The study did seem to
03:29:54have some intriguing,
03:29:55results with progression free survival.
03:29:58And,
03:29:59these patients enrolled generally have
03:30:01these
03:30:02relentlessly
03:30:03progressive meningiomas,
03:30:04and historically, survival is, usually
03:30:07in the six to twelve
03:30:08month range, for these patients
03:30:10who have progressed after radiation
03:30:11and and,
03:30:13surgical therapies.
03:30:14So this seemed intriguing,
03:30:17and, yeah, here's the overall
03:30:18survival.
03:30:19Most patients were censored after
03:30:21a year to a year
03:30:22and a half on on
03:30:23study. We can see more
03:30:24than fifty percent of patients
03:30:25enrolled have survived more than,
03:30:27like, eighteen months. So,
03:30:29that seemed to be pretty
03:30:30good results for this population
03:30:32of patients.
03:30:33The abemaciclib also, had good
03:30:35tolerability. Again, it's a a
03:30:37it's kind of a commonly
03:30:38used drug for breast cancer,
03:30:40and seems to be
03:30:41intriguing for these patients. So,
03:30:43this is gonna be studied
03:30:44in, a larger study to
03:30:46be planned.
03:30:48And then the third study
03:30:49that I'm presenting is, called
03:30:51the METIS study. So it's,
03:30:54an international multicenter phase three
03:30:56study evaluating the efficacy and
03:30:58safety of tumor treating fields
03:30:59or TT fields therapy in
03:31:01patients with non small cell
03:31:02lung cancer brain metastasis.
03:31:05The concept here is to
03:31:06treat patients that have,
03:31:09lung cancer brain METs with
03:31:10tumor treating fields therapy after
03:31:12they received SRS,
03:31:14gamma knife for their brain
03:31:16metastasis. Patients were randomized one
03:31:18to one to receive SRS
03:31:20and then follow-up treatment with
03:31:21the device or best supportive
03:31:23care.
03:31:24Patients would continue therapy on
03:31:26the device past second progression.
03:31:28Patients in the best supportive
03:31:30care, arm were allowed to
03:31:31cross over. Only a small
03:31:33number of people crossed over
03:31:34after second progression of fifteen
03:31:36patients.
03:31:38So the primary endpoint
03:31:40for this study,
03:31:41was the the time to
03:31:42their first intracranial progression
03:31:44or neurological death. There was
03:31:47found to be an advantage,
03:31:50in patients that were randomized
03:31:51into the t t fields
03:31:52arm. Patients had,
03:31:55using the device had,
03:31:57I guess, slower time until
03:31:59intracranial progression,
03:32:00and seemed to overall do
03:32:01better. The depth, between the
03:32:03patients, was kind of similar
03:32:05between the the groups that's
03:32:06down there in the the
03:32:07far corner.
03:32:08The median overall survival,
03:32:10was was no different. Again,
03:32:12a majority of patients died
03:32:13from,
03:32:14systemic,
03:32:15issues related to their cancer,
03:32:17not neurological death.
03:32:19But in using the device,
03:32:21there was overall improvements of
03:32:23quality of life in the
03:32:24patients in the the device
03:32:26arm, and I think that
03:32:27correlates to just delaying the
03:32:29time of neurological progression. So
03:32:30as a neuro oncologist,
03:32:32I wanna delay neurological progression
03:32:34for as long as possible
03:32:36because that progression correlates with
03:32:37neurological disabilities and poor quality
03:32:39of life.
03:32:40So criticisms are brought to
03:32:42this device, you know, making
03:32:44the patients wear the device
03:32:45may decrease their quality of
03:32:46life with having to deal
03:32:47with it, but in fact,
03:32:49when
03:32:50studied and using validated quality
03:32:52of life metrics, quality of
03:32:53life is improved in patients
03:32:54that use the device,
03:32:56simply by virtue of delaying
03:32:58the time until progression.
03:33:00It's another,
03:33:01curve looking at their quality
03:33:02of life
03:33:04outcomes, again, kind of favoring
03:33:06group with the the TT
03:33:07fields.
03:33:08So, in conclusion, the study
03:33:10met its primary endpoint. The
03:33:11time to first intracranial progression
03:33:13was prolonged in patients on
03:33:15the device, compared to best
03:33:16supportive care, and therapy was,
03:33:19was well tolerated.
03:33:21Alright. I made up some
03:33:21time there, doctor Herbst. So
03:33:29You want you want any
03:33:30question?
03:33:31Yeah. Anyone have any questions?
03:33:33How do we make an
03:33:34appointment with you?
03:33:36Oh,
03:33:38well, unfortunately, the survival, for
03:33:41GBM remains grim. It seems
03:33:42to be the worst form
03:33:44of cancer compared to others
03:33:46that have been presented. I
03:33:47mean, it's just remarkable to
03:33:48see all the options. That
03:33:49data for the GU cancers
03:33:51presented at the beginning of
03:33:52the morning, the the survival
03:33:54curves, or what doctor Gold
03:33:55represented with that chart with
03:33:57all the options for long.
03:33:58I I just hope that
03:33:59someday
03:34:00my patients will get those
03:34:02options also. Right. But when
03:34:03patients have brain, when we
03:34:05have patients, we need a
03:34:06consultation. Where do you see
03:34:06your patients? I see patients
03:34:07in,
03:34:08New Haven, Trumbull office. We
03:34:09have another faculty who joined
03:34:09us in June, doctor Sylvia
03:34:10Kurz, and doctor Behring. And
03:34:11we're actively recruiting for another
03:34:12chief to replace doctor O'Muro.
03:34:20So we're trying to boost
03:34:21up
03:34:22the number of docs. And
03:34:24hopefully going along with that,
03:34:25we'll have more treatment options
03:34:26for our patients. Wonderful. Thanks.
03:34:28Thank you.
03:34:33So now, probably in between
03:34:35two surgeries, Elena has come
03:34:37here,
03:34:39to tell us a little
03:34:39bit of an update on
03:34:41gynecologic
03:34:41cancers. Good to see you,
03:34:42Elena.
03:34:44Let's get her slides up
03:34:45here. Nelson, there you go.
03:34:49Thank you so much, doctor
03:34:50Herbst, for including us. We
03:34:52in the world of gonococcal
03:34:54cancer very much like to
03:34:55be included in this, so
03:34:56we greatly appreciate it. I
03:34:58appreciate that I am keeping
03:35:00you from lunch, so I
03:35:02will move nicely.
03:35:03I never wanted like never
03:35:05liked being in the spot.
03:35:07So,
03:35:09I will go through things
03:35:10pretty quickly. There's actually been
03:35:11a lot of very exciting
03:35:13updates
03:35:14in the last year in
03:35:15different conferences.
03:35:18And,
03:35:19this year I have nothing
03:35:21to disclose.
03:35:22This year, unlike years in
03:35:24the past where ovarian cancer
03:35:26is really the only cancer
03:35:28where there is meaningful research,
03:35:30This year, we actually have
03:35:31been very lucky to see
03:35:33advances in many different,
03:35:35ovarian
03:35:36many different gynecologic cancers.
03:35:38So I will start, and
03:35:39I will go very quickly
03:35:40because there's actually a lot
03:35:41of very profound, meaningful work
03:35:44that has improved PFS and
03:35:46OS.
03:35:48And I'll try to just
03:35:49then subdivide it by disease
03:35:51site. So in ovarian cancer,
03:35:52I'm really gonna just mention
03:35:54two and then a couple,
03:35:55subsequently that are a little
03:35:57bit earlier in the development.
03:35:59So the first one is
03:36:00REM two one. This is
03:36:02actually very exciting. This is
03:36:03a study for low grade
03:36:05ovarian cancer.
03:36:07Usually, all ovarian cancer research
03:36:09that's done is in high
03:36:10grade. That's the cancer that's
03:36:11really deadly, but
03:36:13so is low grade. And
03:36:14low grade, traditionally, has been
03:36:16a very surgical disease because
03:36:18chemotherapy is traditional chemotherapy really
03:36:20has not been found to
03:36:21be effective.
03:36:23So REM two one was
03:36:24very exciting.
03:36:26The, previous treatments that we
03:36:28have have overall response rate
03:36:30of zero percent to twenty
03:36:31six percent. So these are
03:36:33both targeted agents.
03:36:34Avelumab is inhibitor of MEK.
03:36:37Dipacamab is the oral inhibitor
03:36:39of FAK,
03:36:40and has showed pretty profound
03:36:42overall response rate,
03:36:44and benefit in PFS and
03:36:47OS.
03:36:48A lot of these p
03:36:49patients additionally have been on
03:36:51MEK inhibitors previously, so this
03:36:53is doubly exciting because MEK
03:36:55inhibitor was the previous really
03:36:56exciting thing that I have,
03:36:58discussed with you that really
03:37:00changed the nature of the
03:37:01treatment of this disease.
03:37:03It continues to be largely
03:37:04surgical disease, but now it's
03:37:06great to see some sort
03:37:07of targeted,
03:37:08personalized options for women.
03:37:11And this is another one
03:37:12that I wanted to select
03:37:14to mention, the ctDNA.
03:37:16In the world of gynecologic
03:37:17oncology over the past few
03:37:18years,
03:37:20we have had very exciting
03:37:21development, of course, of PARP
03:37:22inhibitors. PARP inhibitors are something
03:37:24that we started to use
03:37:26routinely for treatment of ovarian
03:37:27cancer and then subsequently for
03:37:29maintenance. And it really changed
03:37:31the paradigm, the nature of
03:37:32ovarian cancer. You know, women,
03:37:34it became much more of
03:37:36a biomarker driven disease.
03:37:38PFS improved. OS improved. We
03:37:41really appreciated later. It was
03:37:42really mostly so for women
03:37:44with BRCA BRCA, BRCA somaticodermal
03:37:47mutations
03:37:49or homologous recombination deficiency.
03:37:51But we got so excited
03:37:53about it that, unfortunately, we
03:37:55have put all women on
03:37:56it, and then subsequently, this
03:37:57is now the time where
03:37:58a lot of women developed
03:37:59dysplastic syndromes and acute leukemias.
03:38:02And a lot of women,
03:38:03for a number of women,
03:38:05unfortunately, have died from that
03:38:06complication of the PARP.
03:38:08So over the past couple
03:38:09years, we really have struggled.
03:38:10How long do you go
03:38:11on PARP? You know, do
03:38:12we continue how, like, we
03:38:14used to do for five
03:38:15years, or do we stop
03:38:16at two, which is really
03:38:17what is now much more
03:38:18acceptable. And, certainly, what I
03:38:20do now,
03:38:21I try not to go
03:38:22beyond thirty six months.
03:38:24So this is one of
03:38:25the studies about ctDNA,
03:38:27about using that kind of
03:38:29testing,
03:38:30as a biomarker driven testing
03:38:32to determine,
03:38:33whether PARP inhibitor should be
03:38:35continued and for for how
03:38:36long. So this was a
03:38:38very positive study that really
03:38:39was very significant and very
03:38:41positive for being able to
03:38:42detect,
03:38:44the specific mutations,
03:38:46and then was able to
03:38:48help identify which patients should
03:38:50be continued and which one
03:38:51should be stopped.
03:38:54I wanted to mention this
03:38:55one. This is a Yale
03:38:55study, exibapillonevastatin.
03:38:57This is something that we
03:38:57have discovered here. This is
03:38:58something that comes from us,
03:39:00and something we have known
03:39:01for a long time, but
03:39:02really very novel,
03:39:04chemotherapy regimen that's now nationally
03:39:06being tested.
03:39:08This is for a very,
03:39:09very bad prognostic group, a
03:39:10platinum resistant,
03:39:13very, very pretreated ovarian cancer.
03:39:15And, this combination has now
03:39:18been used nationally and has
03:39:19been presented in conferences
03:39:22to show,
03:39:23significant overall response rate of
03:39:25almost forty percent,
03:39:26which is not anything we
03:39:27see in this kind of
03:39:29a very pretreated, poorly prognostic
03:39:31population.
03:39:32And then very briefly, normal
03:39:34therapeutics,
03:39:35multiple different,
03:39:36small molecules that are being
03:39:38studied in ovarian cancer that
03:39:40are all targeted,
03:39:42based and biomarker driven.
03:39:44So individual cancer is actually
03:39:46even more exciting,
03:39:48this year than anything else.
03:39:50I wanna start with really
03:39:51the most exciting one. So
03:39:52I'm sure all of you
03:39:54know about pembrolizumab
03:39:56that has been included in
03:39:57the treatment of the mutual
03:39:59cancer, advanced and poorly prognostic
03:40:02in the mutual cancer. We
03:40:03have started using pembrolizumab,
03:40:05And it was a very
03:40:06interesting and and long journey
03:40:08for which I do not
03:40:08have time to discuss with
03:40:09you this morning. But in
03:40:11short, it was started to
03:40:12be used originally
03:40:14only for,
03:40:16MMR deficient disease.
03:40:18Subsequently,
03:40:20it was FDA approved for
03:40:21MMR deficient disease.
03:40:23And then subsequently, literature has
03:40:25shown that there was overall
03:40:26survival
03:40:27in MMR deficient and proficient
03:40:30tumors, which was actually really
03:40:32incredible because we have not
03:40:33seen this in the world
03:40:34of endometrial cancer, these bad
03:40:36prognostic endometrial cancers, actually, in
03:40:38generations.
03:40:39So very recently, as of
03:40:41August eighth and then August
03:40:43tenth,
03:40:44FDA approved both,
03:40:47dasstolumab
03:40:47that I will mention now
03:40:48and pembrolizumab
03:40:50for use in all patient
03:40:51population with high
03:40:53grade, high
03:40:55prognostic
03:40:56factors, high stage recurrent in
03:40:58the mutual cancer.
03:41:00And the Ruby trial is
03:41:01that the second trial that
03:41:02followed after the additional pembrolizumab
03:41:05about the additional dostrelinab
03:41:07in the same kind of
03:41:08population.
03:41:09And, again, it was the
03:41:11first study in a long
03:41:12time that we have shown
03:41:14very significant,
03:41:15but really profound
03:41:17PFS and then overall survival
03:41:19benefit
03:41:20in women with both,
03:41:22deficient tumors,
03:41:23of course, much more profound
03:41:25benefit, but even proficient tumors.
03:41:27So this is actually something
03:41:29that's literally, like, right off
03:41:31of the press
03:41:33and, like, literally in the
03:41:34middle of the chemotherapy cycles,
03:41:36women now, we we added
03:41:38this. We started adding them
03:41:39directly to their treatment protocol
03:41:41and then for maintenance for
03:41:43two to three years between
03:41:44this and pembrolizumab.
03:41:47This is another,
03:41:49targeted option,
03:41:52GOG thirty forty one,
03:41:54where
03:41:55the similarly addition,
03:41:57of the the valve up
03:41:58to the carboplax or carboantaxel.
03:42:02And this one is even
03:42:03more enticing because there's an
03:42:04addition of a olaparib as
03:42:06pop inhibitor.
03:42:08And there's multiple different groups
03:42:10that show benefit,
03:42:11again, biomarker
03:42:13driven in multiple populations.
03:42:15But super exciting. Like, for
03:42:17the first time, again, a
03:42:18very long time, we're seeing
03:42:19profound PFS and even more
03:42:22OS
03:42:24benefit and really, really exciting
03:42:26that it that it's happening
03:42:27biomarker driven in a really
03:42:29truly personalized care kind of
03:42:31a model.
03:42:33Pembrolizumilovatinib,
03:42:34something that I talked to
03:42:35you last time that was
03:42:36presented at ASCO a few
03:42:37years back.
03:42:38And now, of course, there's
03:42:39multiple studies that's looking. Can
03:42:42can those
03:42:43biomarker driven, these targeted options,
03:42:46can they actually replace
03:42:48chemotherapy? That's, of course, like,
03:42:50the future and the exciting
03:42:51aspect of it. At this
03:42:53point, we did not show
03:42:55that it can. It was
03:42:56not superior
03:42:58to chemotherapy upfront.
03:43:00It was equal,
03:43:02and is being studied further
03:43:04whether it can actually be
03:43:05replacing
03:43:07cytotoxic chemotherapy.
03:43:08Though there is some population,
03:43:10like, women who have had
03:43:11new adjuvant chemotherapy
03:43:13where there was benefit, as
03:43:14you can see, overall survival
03:43:16of thirty four months versus
03:43:17twenty one months.
03:43:19We are so there was
03:43:20some presentation
03:43:21on PARP inhibitors
03:43:23in individual cancers, something that
03:43:24we have suspected for a
03:43:25long time and something that
03:43:27a lot of us use,
03:43:28off trial and off indication.
03:43:31But there was finally some
03:43:33use,
03:43:34presented
03:43:35of PARM inhibitors
03:43:36in the uterine cancers,
03:43:38especially uterus,
03:43:40serious cancers with some benefit.
03:43:42Multiple trials in cervical cancer,
03:43:44I will not really go
03:43:45through them, but, again, targeted
03:43:47options with some,
03:43:49evidence of,
03:43:50benefit.
03:43:51INTERLACE trial, which is a
03:43:53trial of using chemotherapy
03:43:56almost like in a new
03:43:57adjuvant fashion
03:43:58before the chemoradiation
03:44:00in this population
03:44:01that showed some benefit. And
03:44:04certainly a very valid concept
03:44:06and something that's gonna be
03:44:07studied more.
03:44:09More target options in this
03:44:10population.
03:44:13We have studied so there's
03:44:14a pre presentations
03:44:15on nivo and EP,
03:44:17that is better than nivo
03:44:18monotherapy
03:44:19for metastatic ovarian cancer and
03:44:21other extracellular clear carcinomas
03:44:24with sixteen percent CR with
03:44:25this combination,
03:44:27something that we're now using,
03:44:29in practice.
03:44:30There were some surgical trials,
03:44:32and this is something that's
03:44:33been debated for a long
03:44:34time and probably will be
03:44:35debated for many more decades
03:44:36to come, whether there is
03:44:38benefit of lymph node dissection,
03:44:40at the time of surgical
03:44:43debulking. And this is something
03:44:45that no matter, I think,
03:44:46how much we study, old
03:44:47school and new school is
03:44:48gonna do differently. So, nevertheless,
03:44:51evidence shows that there really
03:44:52was no,
03:44:54advance that there was no
03:44:55benefit in this kind of
03:44:57profound surgical debulkings of the
03:44:59lymph nodes.
03:45:02There's a lot of use
03:45:03now and a lot of,
03:45:04entry a lot of,
03:45:06success now in use of
03:45:08myriotuximab
03:45:10in ovarian cancer for folate
03:45:12receptor positive tumors.
03:45:14This has really, again, changed
03:45:15the paradigm and something in
03:45:17receptor tumor in in,
03:45:19platinum resistant disease we use
03:45:21a lot and has used
03:45:22a lot and has been
03:45:23used a lot. So now
03:45:25there were studies that showed
03:45:26pretty similar benefit in in
03:45:28the mutual cancer, which is
03:45:30very exciting.
03:45:32And then more more of
03:45:34the same that I showed
03:45:35to you before, the immunotherapy,
03:45:36the targeted options that is
03:45:38being used now to understand
03:45:39understand which of the patients
03:45:40that are best selected and
03:45:42what is the best time
03:45:43that's best,
03:45:44used in these kind of
03:45:45patient populations and for how
03:45:47long. Thank you so much
03:45:49for including us.
03:45:56Thank you, doctor Ratner. And,
03:45:59thanks for all you do
03:45:59for the center and for
03:46:00our patients.
03:46:01Okay. Next, we have, head
03:46:03and neck cancer,
03:46:05and, one of the other
03:46:06spores that I mentioned earlier
03:46:08this morning. And very happy
03:46:09to have for a a
03:46:10rapid shot,
03:46:12Barbara Burness, to give us
03:46:14a the spore leader to
03:46:15give us a, update.
03:46:20Hi.
03:46:21So when they said rapid
03:46:22shots, I thought it was
03:46:23early for vodka, but I
03:46:24guess it it refers to
03:46:26to gunfire.
03:46:30Okay. So, this was a
03:46:31year, I think, where we
03:46:32finally started to see,
03:46:36the engagement of of many
03:46:37trials looking at novel novel
03:46:40agents in head neck cancer.
03:46:42None that was necessarily practiced
03:46:44to here are my disclosures.
03:46:45None that was necessarily practice
03:46:47changing, but,
03:46:48really showing a lot of,
03:46:50activity. And I think,
03:46:54this we're gonna be having
03:46:55phase three trials that will
03:46:56change things. So,
03:46:58there were three main classes
03:46:59I wanted to talk about,
03:47:00HPV vaccines,
03:47:02ADCs,
03:47:03and targeted therapy, and there
03:47:04were was really a plethora
03:47:06of of,
03:47:07talks. So I just picked
03:47:09one from from each. So
03:47:10among the HPV vaccines, we
03:47:11also saw PDS o one
03:47:13zero one. We also saw,
03:47:15the Q BioPharma compound.
03:47:17I thought that I'd, show
03:47:18you the Hookeepa,
03:47:20h b two two hundred.
03:47:21This is an arenavirus based
03:47:23immunotherapy,
03:47:26together with, pembrolizumab
03:47:28as first line treatment for
03:47:30HPV,
03:47:31positive,
03:47:33head and neck cancer. This
03:47:34was a a thirty eight,
03:47:36patient cohort.
03:47:37They, had minimum follow-up of
03:47:39four and a half months.
03:47:41Among the thirty five evaluable
03:47:43patients, there were thirteen confirmed
03:47:45responses or an objective
03:47:47response rate of thirty seven
03:47:48percent.
03:47:49I think also striking that
03:47:51they saw some CRs. If
03:47:52you look at the, waterfall
03:47:54plot, the CPS
03:47:55greater than twenty patients are
03:47:56shown in blue, so you
03:47:57can see that the,
03:48:00the high response rate was
03:48:01driven by by those patients.
03:48:05But I think that this
03:48:05is something we're we're likely
03:48:07to see move forward as
03:48:08with the q compound as
03:48:10with PDS,
03:48:11o one zero one.
03:48:13A bispecific antibody that got
03:48:15a lot of attention is
03:48:16pedomab.
03:48:17This was given together with
03:48:19pembrolizumab
03:48:20and, presented by Jerome Fayet.
03:48:22So,
03:48:23PETAO is a bispecific that
03:48:25targets eGFR. You all know
03:48:27with cetuximab
03:48:28really are only validated targeted
03:48:30therapy in head neck cancer.
03:48:32And the,
03:48:33other target is LGR five,
03:48:35which is a receptor for
03:48:36Wnt signaling. And Wnt signaling
03:48:38has been associated
03:48:39with EMT, worst prognosis
03:48:41in, head neck cancer.
03:48:43I think we don't totally
03:48:45understand what the characteristics of
03:48:47the binding of this bispecific
03:48:48are when there's not a
03:48:49lot of LGR five around
03:48:51and how that would stack
03:48:52up against,
03:48:53cetuximab.
03:48:54But nonetheless, they took this
03:48:55forward, together with,
03:48:58pembrolizumab.
03:48:59The, total enrolled population was
03:49:01forty five patients, and they
03:49:03only presented results on twenty
03:49:05four.
03:49:06I think that that was
03:49:07due to maturity, not toxicity,
03:49:09but just take it with
03:49:10a little bit of a
03:49:11grain of salt because of
03:49:12that. But you see here
03:49:13the response rate of sixty
03:49:14seven percent, and this is
03:49:16moving forward in phase three
03:49:17studies both in first line
03:49:19and second line. The first
03:49:20line study is pembro with
03:49:21or without, pito.
03:49:23So I think people are
03:49:24very excited about this compound.
03:49:27There were a number of
03:49:28ADCs that were looked at.
03:49:29I pulled out tesodimab vedotin.
03:49:32This is from a presentation
03:49:33by Lova Sun.
03:49:35So,
03:49:37tesodimab vedotin is a tissue
03:49:38factor directed antibody drug conjugate.
03:49:41It's already approved in cervical
03:49:43cancer,
03:49:44and there was a multidisease,
03:49:46study that had a a
03:49:47head neck cohort of, forty
03:49:49patients in it. And you
03:49:50see here confirmed objective response
03:49:52rate of thirty two percent.
03:49:54This is a extremely heavily
03:49:56treated, pretreated population. If you,
03:49:59confine yourself to looking at
03:50:01second and third line patients,
03:50:02forty percent response rate, and
03:50:04some of these extremely durable.
03:50:07Destiny pantumor o two was
03:50:09looking at,
03:50:11TDXD
03:50:12in, patients who had previously
03:50:15treated HER2 expressing solid tumors,
03:50:17and there was a head
03:50:17and neck cohort. This was
03:50:19not actually driven predominantly by
03:50:21mucosal head and neck cancers.
03:50:22This was more salivary gland
03:50:23cancers. But, again, very intriguing,
03:50:25a response rate of nearly
03:50:27forty two percent.
03:50:28This was higher at fifty
03:50:30seven percent for the patients
03:50:31with high HER2 expression by
03:50:33IHC.
03:50:34Didn't seem to be, negatively
03:50:37impacted by prior HER2 therapy
03:50:39or or prior,
03:50:41radiation therapy. And so with
03:50:43the median duration of response
03:50:45of twenty two months, median
03:50:48PFS of twelve months, and
03:50:49OS of twenty three months.
03:50:52This was really very, intriguing.
03:50:56A a phase three study
03:50:57that was presented that I
03:50:58that I think people didn't
03:51:00quite know what to make
03:51:01of, of this, was a
03:51:02comparison
03:51:03of intensity modulated proton therapy
03:51:05with an intentional
03:51:07intensity modulated
03:51:08photon therapy. This is a
03:51:10multicenter study. It accrued over,
03:51:12surprisingly long number of years,
03:51:14but people were very excited
03:51:16about the possibility that, protons
03:51:18could
03:51:19decrease,
03:51:21radiation dose to adjacent normal
03:51:23structures, and it was already
03:51:24known that it would decrease
03:51:26mucositis, adenaphasia,
03:51:28weight loss, and so forth.
03:51:29So this was a randomized
03:51:31study,
03:51:32very heterogeneous,
03:51:34management approaches. So they permitted
03:51:36induction chemotherapy. They permitted both
03:51:38HPV positive and HPV negative
03:51:40cancers. There was stratification
03:51:42on those as well as
03:51:43on smoking status. And then
03:51:45once patients got to the
03:51:46chemo RT,
03:51:47there was a variety of
03:51:48the ways that the chemo
03:51:49was given, but the radiation
03:51:51was seventy gray IMRT or
03:51:53IMPT,
03:51:56and then,
03:51:57restaging
03:51:58and then salvage surgery if
03:52:00if necessary.
03:52:01So,
03:52:03to to the extent that
03:52:04these that these data,
03:52:07can are interpretable,
03:52:09they they looked good for
03:52:10for protons. So the non
03:52:12inferiority p value was zero
03:52:14point zero zero six.
03:52:16And you can see at
03:52:16five years, it looked like
03:52:17maybe there was a five
03:52:18year improvement in progression free
03:52:20survival.
03:52:21And, you can see how
03:52:22long this study has been
03:52:23ongoing with ten year survivals
03:52:25here. But, again, no disadvantage
03:52:28in overall survival and maybe
03:52:29an improvement in overall survival.
03:52:31I think the next slide
03:52:32will show you a little
03:52:33bit why people are skeptical
03:52:34of this trial. So you
03:52:36see in the control arm,
03:52:37forty two percent usage of
03:52:39gastrostomy tubes,
03:52:41not, I think,
03:52:43consonant with contemporary practice where
03:52:45it's probably more fifteen or
03:52:47twenty percent,
03:52:48but it did seem as
03:52:49if IMPT within these centers,
03:52:51reduced that.
03:52:53We showed the long term,
03:52:55follow-up results of thirty three
03:52:56eleven. This is a study
03:52:57you may remember of transoral
03:52:59surgery for low risk HPV
03:53:01associated disease,
03:53:03followed by de intensified postoperative
03:53:05therapy, asking the question of
03:53:07whether or not transoral surgery
03:53:09would would be a strategy
03:53:10for deintensification.
03:53:12Patients were eligible to a
03:53:13t one, t two cancer,
03:53:15and they weren't supposed to
03:53:16have matted nodes.
03:53:19And then, they went through
03:53:20their transoral resection and neck
03:53:22dissection,
03:53:23and then,
03:53:24were categorized as either high
03:53:26risk if they had positive
03:53:27margins, ENE, or five or
03:53:29more nodes. We had hoped
03:53:31really not to have those
03:53:31kinds of patients in the
03:53:32study, but it ended up
03:53:33being about thirty one percent.
03:53:35If they had close margins,
03:53:37minimal ENE, two to four
03:53:39lymph nodes, or PNI LVI,
03:53:41They were called intermediate risk,
03:53:43and those patients were randomized
03:53:44between sixty, which was the
03:53:45old standard,
03:53:46and fifty gray of postoperative
03:53:48therapy. And then patients who
03:53:50had, a single either no
03:53:51nodes or a single node
03:53:52of, three centimeters or less
03:53:54with no ENE,
03:53:55which would have been AJCC
03:53:57seven, n one disease were
03:53:59observed.
03:54:00So we updated this with
03:54:02four and a half year,
03:54:03follow-up. I think, given the,
03:54:05nature of follow-up in the
03:54:06cooperative group, this may be
03:54:07as good as it gets.
03:54:09And, overall,
03:54:10everything continued to look as
03:54:12it had before,
03:54:13overall survival,
03:54:15ninety five percent for the
03:54:17strategy as a as a
03:54:18whole at fifty four months.
03:54:20But we did see, a
03:54:21fall off in progression free
03:54:24survival,
03:54:26for arm a. So we
03:54:28looked at those thirty eight
03:54:29patients a little bit more
03:54:30closely. Eleven of them had
03:54:31n zero disease, and none
03:54:32of those patients recurred. But
03:54:34among the twenty seven patients
03:54:35who had a single involved
03:54:36node with no ENE,
03:54:38the recurrence rate was fourteen
03:54:39point eight percent. That was
03:54:40actually higher than the higher
03:54:42stage patients who got fifty
03:54:43gray. And so and and,
03:54:44interestingly, three of those recurrences
03:54:46were late at at greater
03:54:47than three years. So I
03:54:48think there is a concern
03:54:49that we misunderstood,
03:54:51the ability to deintensify patients
03:54:53with no positive disease, and
03:54:55I think that may be
03:54:56the one kind of practice
03:54:57changing thing from this update.
03:55:00We also looked at, tumor
03:55:01sub site and smoking history
03:55:03and whether or not they
03:55:04predicted outcome
03:55:05and, neither having a base
03:55:07of tongue cancer nor having
03:55:08a smoking history was associated
03:55:10with WARVES outcome. And then
03:55:12lastly, I'll show you a
03:55:13a
03:55:14you know, what we thought
03:55:15was a failed, trial. This
03:55:16was thirty one sixty three.
03:55:18It was a study I'm
03:55:19just about done. It was
03:55:20a study for patients who
03:55:22had nasal or paranasal sinus
03:55:24cancer who were gonna require
03:55:26either orbital generation or resection
03:55:27of the base of skull.
03:55:29And the question was, would
03:55:30induction chemotherapy
03:55:32lead to structure preservation?
03:55:34Naturally, when you put that
03:55:35to patients, they said, well,
03:55:36if there's even a little
03:55:37bit of a chance, I
03:55:38want the chemo. So this
03:55:39study accrued very poorly.
03:55:41But when we looked at
03:55:43the twenty three evaluable patients,
03:55:44structure preservation went from fifteen
03:55:46percent in the control arm
03:55:48to fifty percent in the
03:55:49intervention arm. And if you
03:55:50looked at people who did
03:55:51not have t four b
03:55:52disease, it went from eighteen
03:55:54to seventy one percent, and
03:55:55that was statistically significant. So
03:55:56I think even though this
03:55:58was less than thirty patients,
03:55:59I don't think people are
03:56:00going to be able to
03:56:02run another study that doesn't
03:56:03involve induction. And so I
03:56:04think this does mean that
03:56:06induction is now the standard
03:56:07of care for structure preservation
03:56:09in, nasal and paranasal sinus.
03:56:12Then just
03:56:13lastly,
03:56:14we expected at ESMO
03:56:16to see a phase three
03:56:17study of zapinepant, which is
03:56:19a smack mimetic. This inhibits
03:56:20an inhibitor of apoptosis,
03:56:22potent radiation sensitizer. There had
03:56:24been a completed phase three
03:56:25study, and the first interim
03:56:27analysis was read out as
03:56:28futile. So the abstract was
03:56:30actually pulled from ESMO, but
03:56:32I think,
03:56:33this led to closing down
03:56:34the zapinepant program and is
03:56:36an important update. Okay. Thank
03:56:37you. Sorry.
03:56:43That was great. I I
03:56:44gave Barbara a little latitude.
03:56:45She's built the best head
03:56:46and neck program in the
03:56:46world, and and I I
03:56:48really mean that. Okay.
03:56:50So
03:56:51so our last speaker before
03:56:52lunch, a few announcements. This
03:56:54is gonna be on palliative
03:56:55care with, doctor Elizabeth Prusick,
03:56:58and then the best palliative
03:56:59care is our lunch. Now,
03:57:01when we break for lunch,
03:57:02the faculty, Eddie, who are
03:57:03still here and hopefully most,
03:57:05we're gonna go outside and
03:57:05do a group faculty picture
03:57:07for next year's brochure because
03:57:08we're gonna do this again
03:57:09same time next year. But
03:57:11but now for ten minutes,
03:57:12we're gonna hear from Liz.
03:57:14And by the way, we'll
03:57:14come back from lunch at
03:57:16one o'clock. You can bring
03:57:17your food to the table
03:57:18because we have one of
03:57:19our deans and department chairs,
03:57:21Lucila Ocho Machado, who's gonna
03:57:23tell us about artificial intelligence.
03:57:24Everyone wants to know that.
03:57:25Just so you know what
03:57:26stock to buy. Right? You
03:57:27gotta gotta listen to that.
03:57:28Okay. So so so, Liz,
03:57:29the floor is yours.
03:58:04Apologies for the delay.
03:58:26Just need to go full
03:58:27screen? Or
03:58:29Configuration.
03:58:35There you go. Alright.
03:58:45Alright. Thank you guys for
03:58:47your patience. Thanks for giving
03:58:48me the opportunity to speak
03:58:49with you today. I'm Elizabeth
03:58:50Persich, and I serve as
03:58:51the director of adult and
03:58:52patient and palliative,
03:58:53care consult at Yale New
03:58:55Haven Hospital and also the
03:58:56firm chief for medical oncology.
03:58:57I'm trained in both, and
03:58:58I I focus my care
03:58:59in the,
03:59:00care of our admitted patients.
03:59:02I have no financial disclosures.
03:59:04The annual meeting this year
03:59:05focused on comfort to cure.
03:59:07We as cancer physicians, we
03:59:08all aspire to cure. However,
03:59:10cancer remains the lead the
03:59:11second leading cause of death
03:59:12in the United States, and
03:59:13the treatments that we offer
03:59:14has significant physical, social, emotional,
03:59:17financial, and time toxicities.
03:59:19All of our patients, regardless
03:59:20of age, stage, and treatment
03:59:21intent who face advanced cancer,
03:59:23will benefit from the integration
03:59:25of palliative care throughout their
03:59:26disease course. I'm gonna focus
03:59:28today on palliative care delivery
03:59:29and also symptom management. And
03:59:31in the interest of time,
03:59:32I'll just dive right in.
03:59:33I'm gonna begin with this
03:59:34study, highlighting the early palliative
03:59:37care among patients diagnosed with
03:59:38advanced cancers between twenty ten
03:59:40and twenty nineteen.
03:59:41As we know, ASCO recommends
03:59:43the early integration of palliative
03:59:45care, throughout the or from
03:59:47sorry. Early integration for all
03:59:49patients with advanced cancers. However,
03:59:51evidence of its use in
03:59:52the role of practice patterns
03:59:53and organizational characteristics in its
03:59:55implementation is limited.
03:59:57This study looked at SEER
03:59:58Medicare data of patients sixty
04:00:01five and older with newly
04:00:03diagnosed advanced cancers in between
04:00:05twenty ten and twenty nineteen
04:00:06with six months or greater
04:00:07survival.
04:00:08Early palliative care was identified
04:00:10by claims
04:00:11within ninety days of diagnosis
04:00:13or up to their first
04:00:14hospice admission.
04:00:16And this study looked over
04:00:17the percent of patients receiving
04:00:18palliative care each year. As
04:00:20you can see and as
04:00:21we all know, palliative care
04:00:22integration has increased significantly
04:00:24over the past decade or
04:00:26more. It went from just
04:00:27one percent of patients in
04:00:28twenty ten to over ten
04:00:29percent in twenty nineteen.
04:00:32There were statistically significant difference
04:00:34or sick it was statistically
04:00:36significant across disease types as
04:00:38you can see above.
04:00:40There were significant variations in
04:00:42patient factors as well as
04:00:44clinician and organizational factors. So
04:00:46patients who had higher comorbidities,
04:00:49older age, females more than
04:00:50males,
04:00:51nonhispanic
04:00:52black patients, and those residing
04:00:54in metropolitan areas were more
04:00:55likely to access early palliative
04:00:57care throughout this period. For
04:00:59nonpatient factors, female clinicians,
04:01:01those of younger ages, and
04:01:03also larger organizations
04:01:04with integrated palliative care services,
04:01:06unsurprisingly, were associated with more
04:01:08early integration.
04:01:09What I found interesting here
04:01:11was that the variation in
04:01:13terms of early palliative care
04:01:14use was more dependent on
04:01:16non patient factors rather than
04:01:18individual patient factors.
04:01:22More patients with advanced cancer
04:01:24are receiving early palliative care,
04:01:25but utilization overall remains low
04:01:27with only ten percent of
04:01:28patients nationally as of twenty
04:01:30nineteen receiving early palliative care,
04:01:31and there were significant variation
04:01:33by clinician and organizational characteristics.
04:01:38The takeaway here is that
04:01:39early palliative care has been
04:01:40increasing over time, but the
04:01:41differences in terms of receipt
04:01:42is really more about us
04:01:43as clinicians and organizations
04:01:45rather than our patients and
04:01:46their individual
04:01:48needs.
04:01:50Second, I'm gonna highlight a
04:01:51multisite randomized trial of sept
04:01:54palliative care for patients with
04:01:56lung cancer. This is a
04:01:58study led by Jennifer Temel,
04:02:00out of Mass General.
04:02:01So we all know that
04:02:02early integration of palliative care
04:02:04is important from the time
04:02:05of diagnosis. It can aid
04:02:06both patients and their caregivers.
04:02:08However, this model has not
04:02:09been widely implemented as we
04:02:10saw in our previous study
04:02:12given the shortage of palliative
04:02:13care clinicians and also challenges
04:02:15providing palliative care visits throughout
04:02:17the course of cancer treatment.
04:02:19So this study looked to
04:02:21how we can deliver more
04:02:22patient centered and less resource
04:02:23intensive palliative care
04:02:25looking at a stepped palliative
04:02:27care model, which I'll outline
04:02:28here. So this graphic outlines
04:02:30the step palliative care approach
04:02:31where palliative care visits are
04:02:32stepped up in response to
04:02:33worsen quality of life, rest
04:02:35measured by the fact that
04:02:36the study looked at patients
04:02:37with advanced lung cancer. I
04:02:39think, realistically, this is how
04:02:41many more of us practice
04:02:42responding to patient need rather
04:02:44than automatic referrals at the
04:02:45time of advanced cancer.
04:02:48So patients with advanced lung
04:02:49cancer here were diagnosed within
04:02:51who were diagnosed within twelve
04:02:52weeks with a performance status
04:02:53of zero zero to two
04:02:55were randomized to either stepped
04:02:56palliative care,
04:02:58or early palliative care between
04:02:59twenty eighteen and twenty twenty
04:03:00two. Patients with an initial
04:03:02palliative care visit within four
04:03:04weeks of enrollment had regularly
04:03:05scheduled quality of life assessments
04:03:07measured by a fact l.
04:03:09Patients randomized to early pallet
04:03:12I'm sorry. Patients randomized to
04:03:13early palliative care had visits
04:03:14every four weeks rather than
04:03:15in response to decreased quality
04:03:17of life per fact l
04:03:18on a stepped group. And
04:03:19the primary aim of this
04:03:20study was to look at
04:03:21non inferiority for quality of
04:03:23life as measured by the
04:03:24fact l at twenty four
04:03:25weeks.
04:03:28So what did they find?
04:03:30The mean number of palliative
04:03:31care visits at twenty four
04:03:32weeks was lower for the
04:03:33stepped palliative care group compared
04:03:36to the early palliative care
04:03:37group, which is not surprising
04:03:39and probably beneficial given resource
04:03:41limitations on the palliative care
04:03:42clinician side.
04:03:44Importantly, quality of life scores
04:03:45at week twenty four for
04:03:47those assigned to stepped versus
04:03:48early palliative care were noninferior.
04:03:51The rate of end of
04:03:52life communication was similarly noninferior
04:03:54between the stepped palliative care
04:03:56and the early palliative care
04:03:57group.
04:03:58However,
04:03:58noninferiority
04:03:59was not demonstrated
04:04:01for days on hospice.
04:04:04Stepped palliative care approaches provided
04:04:06at the time of worsening
04:04:07functional status is not inferior
04:04:08to early palliative care in
04:04:10terms of quality of life
04:04:11or end of life communication
04:04:12at twenty four weeks.
04:04:14However, step palliative care was
04:04:15associated with fewer days of
04:04:16hospice enrollment at the end
04:04:18of life.
04:04:20This next study, I'll go
04:04:21over briefly in the interest
04:04:22of time, but I did
04:04:24wanna touch upon this this
04:04:25work.
04:04:27This is a study looking
04:04:28at behavioral economics and analytics
04:04:30to improve palliative care and
04:04:31advanced cancer.
04:04:32It used an algorithm algorithm
04:04:34based palliative
04:04:35algorithm based,
04:04:37approach to identify patients at
04:04:39highest risk for palliative care
04:04:40referrals.
04:04:41So early specialist palliative care
04:04:43can improve these outcomes for
04:04:44patients with advanced solid tumor
04:04:46malignancies, but most patients don't
04:04:47receive a palliative care referral
04:04:49before death. As we saw,
04:04:50ninety percent of patients in
04:04:51twenty nineteen did not receive
04:04:53any, palliative care whatsoever.
04:04:55Barriers to palliative care may
04:04:57include
04:04:58difficulty identifying high risk patients
04:04:59and also physician referral burden,
04:05:01and that is what this
04:05:02study aimed to address.
04:05:03It was a two arms,
04:05:05pragmatic cluster randomized clinical trial
04:05:08that looked at patients with
04:05:09advanced
04:05:10lung cancer or noncolorectal
04:05:12GI cancers
04:05:13using automated algorithm
04:05:16embedded in the EHR that
04:05:17assigned each patient a risk
04:05:18score.
04:05:20Patients in the intervention arm
04:05:22on the left received,
04:05:23their oncologist received weekly default
04:05:25EHR notifications that prompted specialty
04:05:28palliative care referral. If oncologists
04:05:30did not opt out,
04:05:32a nursing coordinator reached out
04:05:33to the patient, introduced palliative
04:05:35care, and offered to set
04:05:35up a visit. In the
04:05:36control arm on the right,
04:05:38oncologist referred at their discretion
04:05:40as per
04:05:41usual.
04:05:43They use behavioral economic methods,
04:05:45including normative and risk based
04:05:47messaging with explanation.
04:05:48What I found interesting here
04:05:50was that in their opt
04:05:51out message to clinician, it
04:05:52was also assigned by the
04:05:53chief medical officer and the
04:05:54president, which I of the
04:05:56organization, which I think probably
04:05:57helped, with adherence as well.
04:05:59So, in this study, there
04:06:02were five hundred and six
04:06:04two patients,
04:06:05two hundred and ninety six
04:06:06in the intervention arm. The
04:06:07risk score in terms of,
04:06:09severity of illness and comorbidities
04:06:10was equivalent between the groups.
04:06:12In the intervention arm, nearly
04:06:13ninety percent of clinicians did
04:06:14allow for palliative care referrals.
04:06:16There was no opting out.
04:06:17And of the patients contacted,
04:06:18eighty percent,
04:06:20chose to schedule palliative care
04:06:22visits. As you can see
04:06:23in this graphic, unsurprisingly, the
04:06:25intervention group had, significantly higher
04:06:27rates of completed palliative care
04:06:29visits with an adjusted odds
04:06:30ratio of five point four.
04:06:33It did result in decreased
04:06:35chemotherapy
04:06:36at the end of life.
04:06:37So of patients in the
04:06:38intervention arm, it was six
04:06:39point five percent versus sixteen
04:06:40percent of the control group.
04:06:41There was no difference in
04:06:42quality of life or hospice
04:06:44use among the decedents.
04:06:47I'm going over things quickly
04:06:49because I know I'm standing
04:06:49between you and lunch.
04:06:52But, in conclusion, there is
04:06:53this is a large community
04:06:55oncology network algorithm based default
04:06:57palliative care pro referrals were
04:06:58acceptable to clinicians and led
04:07:00to a a threefold increase
04:07:01in specialty palliative care referral
04:07:03and completion as well as
04:07:04decreased end of life,
04:07:05chemotherapy.
04:07:07This approach was supported by
04:07:08physicians in terms of the
04:07:10criteria for selection in those
04:07:12process and did was associated
04:07:14with decreased end of life
04:07:15chemotherapy.
04:07:15We all work so hard
04:07:17for our patients. I think
04:07:18this study reminds us that
04:07:19we can also use data
04:07:20analytics and,
04:07:21AI to help us work
04:07:23smarter too.
04:07:26This study I wanted to
04:07:27touch upon briefly. I've got
04:07:29two more to go.
04:07:30But what I wanted to
04:07:31highlight here was,
04:07:33the effectiveness of early palliative
04:07:34care delivered via telehealth
04:07:36versus in person.
04:07:38Clearly, accessing palliative care, referring
04:07:40to palliative care is a
04:07:41challenge, and our patients who
04:07:42benefit the most from this
04:07:43with advanced cancer often have
04:07:45significant comorbidities in terms of,
04:07:47mobility,
04:07:48time toxicity, and just a
04:07:50lot of interface with health
04:07:51care system. So how can
04:07:52we help overcome these barriers?
04:07:54This study looked at over
04:07:56twelve hundred patients with advanced
04:07:57non small cell lung cancer
04:07:58diagnosed within twelve weeks who
04:08:00are enrolled across twenty two
04:08:01cancer centers. This was done
04:08:02between twenty eighteen and twenty
04:08:04twenty three.
04:08:05They were randomly assigned to
04:08:07meet with a palliative care
04:08:08clinician every four weeks,
04:08:10through the course, either by
04:08:12video or tele video or
04:08:14in person visit, and they
04:08:15had self reported quality of
04:08:17life measures, at weeks twelve
04:08:18and twenty four. The primary
04:08:20aim was to evaluate the
04:08:21equivalence of telehealth versus in
04:08:23person at quality of life
04:08:24at twenty four weeks measured
04:08:25by Factel.
04:08:27Nearly eighty percent of caregivers
04:08:28were also enrolled in this
04:08:29study, which I think was
04:08:30really important as well.
04:08:33Quality of life for patients
04:08:34assigned to the telehealth group
04:08:36was equivalent to those receiving
04:08:37in person early palliative care,
04:08:39which was significant.
04:08:41And review of documented topics
04:08:43that were discussed in the
04:08:44palliative care visit were also
04:08:45similar between groups.
04:08:47So exploratory
04:08:48outcomes also demonstrated that there
04:08:50was no difference in terms
04:08:51of depression or anxiety symptom
04:08:52between groups, which I think
04:08:53is also reassuring.
04:08:56So no difference in a
04:08:57range of patient reported outcomes
04:08:59for this. More caregivers did
04:09:00attend in person versus video
04:09:02visits, and there was no
04:09:03difference in terms of quality
04:09:04of life between the, in
04:09:06person and telehealth visits. So
04:09:07I think this is another
04:09:08important tool to expand access
04:09:10to palliative care and gives
04:09:11us reassurance that in many
04:09:13respects, the, visit visits will
04:09:16be equivalent in terms of
04:09:17effect. Looks like you're wrapping
04:09:18me up, so I will
04:09:19skip over the rest.
04:09:22But I will I do
04:09:23wanna touch on two things
04:09:25quickly. Yes. I will give
04:09:27you another minute.
04:09:29Sounds good. Palliative.
04:09:31Yes.
04:09:34Yeah. I will go a
04:09:35one liner for each. But
04:09:37That came. Start the line.
04:09:39As soon as we stop.
04:09:41One study I wanted to
04:09:42touch upon, comprehensive prescription drug
04:09:44monitoring mandates. They help us
04:09:46keep our patients safe with
04:09:47the opioid epidemic and all
04:09:48the, opioids that we do
04:09:50prescribe for, advanced cancer. What
04:09:52I wanted to highlight here
04:09:53is that there are significant
04:09:54disparities in terms of how
04:09:56these mandates
04:09:57affect our patients.
04:10:00There is up to a
04:10:01ten percent decrease in opioid
04:10:02prescriptions for patients with advanced
04:10:04cancer nearing the end of
04:10:05life, and Stacy said did
04:10:06implement the PDMP programs and
04:10:08that this disproportionately
04:10:09non Hispanic black patients. So
04:10:10I think we all
04:10:12need to be mindful of
04:10:12health equity issues when it
04:10:13comes to cancer care, especially
04:10:14at end of life for
04:10:14all of our patients. And
04:10:14secondarily, this one got me
04:10:15pretty amped up before lunchtime,
04:10:17but payer
04:10:18denial for prior authorization request
04:10:19for
04:10:24long acting pain medications, a
04:10:26major issue we all know.
04:10:28Ninety percent of prior authorizations
04:10:30for patients with advanced cancer
04:10:31and severe cancer associated pain,
04:10:34are ultimately approved. The average
04:10:36length of time it takes
04:10:37to deal with these approvals
04:10:39is two days, And uppatients
04:10:41that do have denials, which
04:10:42is about five percent, they
04:10:43get their medications completely denied.
04:10:45A quarter of those patients
04:10:46will end up requiring acute
04:10:47care either in the emergency
04:10:49room or in the hospital,
04:10:51and one in ten of
04:10:52those patients will pay out
04:10:53of pocket. So I think
04:10:54this tells us we all
04:10:55need to advocate for improved
04:10:57transplant prior authorization processes. And
04:10:59in the meantime, make sure
04:11:00we give at least three
04:11:01to five days to address
04:11:02these issues so our patients
04:11:03aren't suffering unnecessarily at the
04:11:05end of life.
04:11:07Palliative care has been growing,
04:11:08but not enough. There are
04:11:09a lot of care delivery
04:11:10innovations that can increase appropriate
04:11:12referrals and decrease clinical burden
04:11:15for our oncologists and improve
04:11:16staffing,
04:11:17and opioid state insurance policies
04:11:19may have unintended consequences that
04:11:21can be detrimental to our
04:11:22patients.
04:11:26A palatal exam.
04:11:28Thanks.
04:11:32Okay. That was great. TPN
04:11:34is in the other room.
04:11:34Okay. So now
04:11:37so now we we have
04:11:38lunch.
04:11:39We are gonna start the
04:11:41lecture at one o'clock. We
04:11:42have the tables. You can
04:11:43bring the lunch back to
04:11:44your tables and eat.
04:11:46There's plenty of food here.
04:11:47Everyone will get food. There's
04:11:49also exhibits and so forth
04:11:51in the other room, which
04:11:52I urge people to visit.
04:11:53So those who are closest
04:11:54to the food, you picked
04:11:55a good seat.
04:14:38Okay, everyone. We're gonna start
04:14:40in one minute.
04:14:41One minute.
04:15:08This is it. It seats
04:15:08my profile. Yeah.
04:15:19Okay.
04:15:21Let's let's I'll get to
04:15:22our seats. I
04:15:23this lecture will be better
04:15:24if it's quiet.
04:15:27Okay. I'm gonna give it
04:15:28another thirty seconds.
04:15:30Good to see all our
04:15:31fellows back. I went out
04:15:32and got the. Yes.
04:15:35They need to hear them.
04:15:40You know Scott? Have you
04:15:41ever met Scott? Have you
04:15:43met? Yes. Have you met?
04:15:47He's in he does a
04:15:48lot of our quality work
04:15:49and outcomes, and
04:15:53he needs someone to get
04:15:54to know me.
04:16:02Okay.
04:16:13Okay.
04:16:24Thank you all. It's been
04:16:26a a wonderful day,
04:16:28and and the day still
04:16:29has a lot to go.
04:16:30We,
04:16:31I'm I'm really excited. We
04:16:32have a whole big section
04:16:33on hematology. Right, doctor Huntington?
04:16:36I see doctor Helene and
04:16:37doctor Podostev here.
04:16:39I I got a lot
04:16:40to learn. You know?
04:16:41Okay. So so now so
04:16:43now what we're gonna do
04:16:44is we have a special
04:16:45keynote lecture. And as has
04:16:47been our tradition the last,
04:16:48I guess, thirteen years that
04:16:50I've been doing this, is
04:16:51we wanna bring someone in
04:16:52who's gonna raise the bar.
04:16:54You You know, what's the
04:16:54future of health care of
04:16:56of of medical research? And
04:16:58and we really were so
04:16:59fortunate this year that our
04:17:01committee,
04:17:02Pam Kuntz,
04:17:03Sue Evans,
04:17:05Kiran Taraga,
04:17:06myself,
04:17:08Scott Huntington.
04:17:09We our first choice agreed
04:17:10to come, and that's doctor
04:17:12Lucila Ocho Machado, who's who's
04:17:14here on my right. She's
04:17:15the Waldamer Vars Zenowitz professor
04:17:17of medicine and bioinformatics
04:17:19bioinformatics
04:17:21and data science, deputy dean
04:17:23for biomedical informatics,
04:17:24and now there's a chair
04:17:26of a new department of
04:17:27biomedical informatics and data science.
04:17:29Her CV is enormous, so
04:17:31I'll just give a few
04:17:32highlights. She's been one of
04:17:33the leaders in using big
04:17:35data,
04:17:36to solve health care issues.
04:17:39She,
04:17:40has been in a number
04:17:41of, amazing places, the Brigham
04:17:44and Women's Hospital,
04:17:45MIT,
04:17:47out in California and San
04:17:48Diego for many years, where
04:17:50she,
04:17:51linked, the databases for many
04:17:52of the hospitals in the
04:17:53California system. I think she'll
04:17:54probably tell us about how
04:17:56she did some of that
04:17:57or certainly,
04:17:58things that came from that
04:18:00and, and really has been
04:18:01a leader in how to
04:18:02use big data. And, I
04:18:03know that one of the
04:18:04initiatives in the last two
04:18:05or three years here at
04:18:06Yale has been to build
04:18:07that, so we're very happy
04:18:08that she's joined us and
04:18:09been recruited. She has many
04:18:11different awards, including election to
04:18:12the National Academy of Medicine,
04:18:14one of the highest honors.
04:18:15So it's really wonderful to
04:18:16have you here, and thank
04:18:17you for coming to our
04:18:19ASCO review and
04:18:25beyond. Thanks for the invite
04:18:27and for the opportunity to
04:18:28talk a little bit about
04:18:29AI, which, excites a lot
04:18:31of people, scares a few
04:18:33others,
04:18:34or,
04:18:35both.
04:18:36So,
04:18:37I also thanks for who
04:18:39whoever,
04:18:40came up with the title
04:18:41because my typical is challenges
04:18:43and opportunities, but this is
04:18:45much more exciting,
04:18:47I would think. So I'll
04:18:49I'll do an overview. So
04:18:50for for those who know
04:18:51AI very much, it's a
04:18:52good time to get dessert
04:18:54or coffee and so on.
04:18:56For those who don't and
04:18:57are interested in learn, I
04:18:59will go a little bit
04:19:00about the basics. So I'll
04:19:02I'll, simplify.
04:19:03Pardon me for simplifying a
04:19:05lot of things, but I
04:19:06think it helps to understand
04:19:07what the issues are. So
04:19:09here's an overview, and I'll
04:19:11I'll get to talk about
04:19:12large language models, whichever one,
04:19:15is a it's an obligatory
04:19:17part of
04:19:18the presentation. But, AI has
04:19:20been around for quite a
04:19:21while. In the eighties was,
04:19:24expert systems. In the nineties
04:19:26was neural networks,
04:19:28other machine learning,
04:19:30methods, and and,
04:19:32past decade was deep learning.
04:19:34And now we are in
04:19:36the era of generative AI.
04:19:37So I will review a
04:19:38little bit of what that
04:19:40is, and the history will
04:19:42be written for you, at
04:19:43the thirties and forties what
04:19:45AI is going to do.
04:19:47So as you know, the
04:19:48New England Journal created a
04:19:50whole new, cascade journal called
04:19:52the AI in medicine.
04:19:55The JAMA
04:19:56network of journals has a
04:19:58special channel
04:19:59on AEI because the volume
04:20:01of publications and research in
04:20:03this area has been
04:20:05increasing,
04:20:06exponentially.
04:20:07But I will go back
04:20:09to a time where most
04:20:10of you weren't even born.
04:20:12In nineteen sixty six, the
04:20:14first chatbot
04:20:16was,
04:20:18the publication was out. And
04:20:20this, as you look on
04:20:21the right hand side, is
04:20:22at a time where the
04:20:23computer terminals didn't even have
04:20:25a monitor.
04:20:26It was all,
04:20:29printed
04:20:30as, the person interacted with
04:20:32the computer. An emulator exists.
04:20:34So if you Google ELISA
04:20:36emulator,
04:20:37you will be able to
04:20:39talk
04:20:39to a computer in the
04:20:41sixties, and it might,
04:20:43prove very entertaining
04:20:45as it is today to
04:20:46talk to chat GPT.
04:20:48So ex persistence came about
04:20:50in the eighties, and they
04:20:52were specific to diagnosing particular
04:20:55conditions such as the one
04:20:56described here.
04:20:57And they were rule based.
04:20:59They will elicit what rules
04:21:01a particular clinician uses to
04:21:03diagnose glaucoma and treat infectious
04:21:05diseases
04:21:06and so on. And, at
04:21:08that time, there wasn't enough
04:21:10data. There wasn't a monitor.
04:21:11As you can see, this
04:21:12is a a terminal
04:21:14for what's called a mini
04:21:15computer
04:21:17at the same time, and
04:21:18the mini was very big
04:21:19as you can tell.
04:21:21So way less,
04:21:23powerful than than any of
04:21:25our,
04:21:27phones.
04:21:28So I'll just go back
04:21:29to the the basics here.
04:21:31What is machine learning doing,
04:21:33and how does it do
04:21:34it? So think about inputs
04:21:36to a system, to an
04:21:39equation, for example, in which
04:21:40you have the age of
04:21:41patient's thirty four, variable one
04:21:43value is four, and you
04:21:45have a series of values.
04:21:46And then you are predicting
04:21:48some outcome, such as development
04:21:51of disease or,
04:21:54metastases and and so on.
04:21:55So with a binary outcome,
04:21:57yes or no,
04:21:59could be,
04:22:00death or or alive. And
04:22:02all that happens is this
04:22:04mathematical
04:22:05equation multiplies the inputs by
04:22:08some coefficients
04:22:10and then passes through a
04:22:11function. In this case, it
04:22:13was called a sigmoid or
04:22:14logistic function and predicts
04:22:17what the estimate would be.
04:22:18So forty percent chance of,
04:22:22death, for example, in this
04:22:24example here. And on the,
04:22:27left hand side, you see
04:22:28how it's depicted as a
04:22:30two d
04:22:32graph. Right? So the individuals
04:22:35in, green might be the
04:22:37ones who survive, whereas the
04:22:39other ones in
04:22:42red, the ones who did
04:22:43not survive. And all this
04:22:45equation is doing is drawing
04:22:46a line and saying from
04:22:48this side, they will likely
04:22:50to survive with this probability
04:22:52for each one of the
04:22:53points.
04:22:54So, again, this is a
04:22:55regression model way before
04:22:58any AI was around. So
04:23:00I'll contrast that, in a
04:23:02minute. So just remember a
04:23:04few things about logistic regression.
04:23:06No one is
04:23:07scared of it in the
04:23:09first place. We know how
04:23:10it works.
04:23:11It's reproducible.
04:23:12It's good for interpretation
04:23:14because it will say if
04:23:15you've standardized the inputs, then
04:23:17the the value of the
04:23:18coefficient will tell, oh, this
04:23:20is important. This is indicative
04:23:22of a higher probability of
04:23:24death or lower. So the
04:23:26risk factors, you would know
04:23:28what they mean.
04:23:30And then as I said,
04:23:32very few parameters, easy to
04:23:34compute.
04:23:35Calculators
04:23:36of the past would do
04:23:37it. So so it's very,
04:23:39very computer
04:23:41nonintensive
04:23:42and reproducible.
04:23:43And on the right hand
04:23:44side, you see the line
04:23:45that separates this class.
04:23:48Now think about machine learning,
04:23:50and all it does is
04:23:52a more complex,
04:23:54model that includes
04:23:55one or more hidden layers
04:23:57in there, but the
04:24:00principle is the same. Instead
04:24:02of coefficients, these are called
04:24:04weights,
04:24:05and there are multiplications
04:24:07that will happen. And then
04:24:08we'll finally,
04:24:10issue an estimate of the
04:24:12probability of the disease. So
04:24:14very
04:24:14similar. However, on the right
04:24:16hand side, you see that,
04:24:18separating
04:24:20a line can be curved,
04:24:21can be much more complex
04:24:23than the other one. Right?
04:24:25So,
04:24:26the trade off is is
04:24:27more complex to compute,
04:24:30but it separates
04:24:31the cases better as you
04:24:33can see here.
04:24:34But it's not so good
04:24:35for interpretation
04:24:36because you no longer has
04:24:38have those coefficients that you
04:24:39can say, oh, this is
04:24:40good. This is bad,
04:24:42and so on.
04:24:43It works well for most
04:24:45problems
04:24:46if lots of data are
04:24:47available. So it's much more
04:24:49data hungry than others. Many
04:24:51parameters to compute, and it's
04:24:53hard to to compute, but
04:24:55it's also reproducible.
04:24:56The moment you freeze the
04:24:58training and you say these
04:25:00are the weight, it's completely
04:25:02reproducible
04:25:03model. So that's one of
04:25:04the myths that needs to
04:25:06be dissipated
04:25:07out there. Of course, if
04:25:08you keep updating, but it's
04:25:10just like a regression model.
04:25:12If you kept updating, of
04:25:13course, it would change. But
04:25:15but,
04:25:16they behave the same way.
04:25:18So this is the old
04:25:19machine learning. Right? The deep
04:25:21learning,
04:25:22what happens, computers got better.
04:25:24It added a whole lot
04:25:25of hidden layers, and that's
04:25:27why it's called deep.
04:25:28And it can do this
04:25:29classification
04:25:30problems that that I'm showing
04:25:32up here. So this is
04:25:34kind of old news. It
04:25:35was there for a long
04:25:36time. And later, we'll talk
04:25:37about the generative models.
04:25:40Regression and machine learning models
04:25:41have been in medicine for
04:25:42a long time. So a
04:25:44lot of the issues that
04:25:45are surfacing now
04:25:47apply to to this,
04:25:49other models except no one
04:25:51talked about them much.
04:25:53And the models are only
04:25:55as good as the data
04:25:56that that would was used
04:25:58to estimate those parameters. Right?
04:26:01And in that, you can
04:26:02tell, you know, if you
04:26:03bias the data, you'll bias
04:26:05the model. There's no secret
04:26:08about that. So,
04:26:10you might not know, but,
04:26:11Framingham Cardiovascular
04:26:13Risk Model is a regression
04:26:15model.
04:26:16Same as for the model
04:26:17for end end stage liver
04:26:19disease, which in fact,
04:26:21ranks people for the liver
04:26:23transplant provided there is a
04:26:25match. So very influential
04:26:27model would put people with
04:26:29higher risk at the top
04:26:31of the queue,
04:26:32for the liver transplant.
04:26:34And there are many, many
04:26:35other,
04:26:36regression models out there and
04:26:38some machine learning models
04:26:40as well, all doing this,
04:26:42classification
04:26:43trap a task that I
04:26:45I talked about. So what
04:26:47are health care organizations,
04:26:49want to use AI for?
04:26:51Increasing efficiency.
04:26:53There is this
04:26:55pilot of Ambien listening now
04:26:57that will summarize your notes
04:26:59having heard what happens during
04:27:01the visit.
04:27:03It there is another,
04:27:05application very common,
04:27:07becoming more common now, which
04:27:09is to generate the draft
04:27:11draft replies for your inbox
04:27:13as a clinician.
04:27:15So those things are happening
04:27:16right now.
04:27:18We also want to decrease
04:27:20errors. Right? Reduce diagnostic
04:27:22errors, therapy selection errors, promote
04:27:25safety,
04:27:26do reminders for differential diagnosis,
04:27:29and so on. So there's
04:27:30a lot coming up, but
04:27:31the first applications will be
04:27:33very much,
04:27:35around the efficiency,
04:27:37around,
04:27:38billing.
04:27:39You know, have the all
04:27:40the billing codes been,
04:27:42included and and things like
04:27:44that.
04:27:45Now at Yale, there is
04:27:46a lot happening in machine
04:27:48learning.
04:27:49Area here is a slide
04:27:50from, cardiologist,
04:27:52Rohan Kara,
04:27:54in which he developed
04:27:55an app that you can
04:27:57anywhere in the world, if
04:27:58you can, take a picture
04:28:00of the twelve lead EKG,
04:28:02it won't interpret it for
04:28:04you. So very exciting application
04:28:07as well as integrating data
04:28:09from wearables. So that this
04:28:11is happening
04:28:12at Yale right now.
04:28:14Of course, your colleague, Sanjay
04:28:15Anisha,
04:28:17is working on,
04:28:18other, like, imaging based biomarkers
04:28:21for lung cancer. So imaging
04:28:22is an area where AI
04:28:24is is being a lot
04:28:26used.
04:28:27And in this case, he
04:28:28wants to
04:28:30determine whether through through imaging
04:28:33you can start to predict
04:28:35what therapy works best for
04:28:37a particular,
04:28:39type of patient. So exciting
04:28:41applications.
04:28:42In practical terms, there are
04:28:44a lot of AI,
04:28:47software that that's been,
04:28:49utilized in the, radiology department,
04:28:53that recognize
04:28:55images that
04:28:57tries to understand the disease
04:28:59progression
04:29:00and so on. So there
04:29:01was a lot happening. It's
04:29:02still piecemeal in,
04:29:04several areas.
04:29:06It's some of it is
04:29:08still research, but there are
04:29:09a lot of applications already,
04:29:12happening.
04:29:13Now we are very interested
04:29:14in polygenic risk scores, which
04:29:17are an equation of the
04:29:18type that I shown before.
04:29:20It's a regression
04:29:22equation, but it has a
04:29:23ton of, parameters. Right? It
04:29:25has a ton of inputs,
04:29:27which are basically the mutations
04:29:29in genes
04:29:30in in other,
04:29:31ancestry variations.
04:29:34We are cognizant that it
04:29:36could introduce biases
04:29:37to that lead to discrimination.
04:29:39So we have to be
04:29:40very, very careful,
04:29:42particularly because the training data,
04:29:44the data that are available
04:29:45for the models,
04:29:47typically comes from,
04:29:49European populations,
04:29:50and and the majority. In
04:29:52other populations, these are not
04:29:54as,
04:29:54well included.
04:29:56There's also the issue of
04:29:58privacy
04:29:59and,
04:30:01the fact that genetic studies
04:30:03typically ignore
04:30:04individuals
04:30:05in which the mixed ancestry
04:30:07is above a certain certain
04:30:09threshold. So that that is,
04:30:11actually very sad because it
04:30:13leaves out,
04:30:14a lot of potential participants
04:30:16in these studies.
04:30:18So,
04:30:18like,
04:30:19trying to represent here that
04:30:21a trait is a function
04:30:23of several,
04:30:25determinants that can be genetics,
04:30:28mutation, variant effects, or ancestral
04:30:31effects, and also social determinants
04:30:33and exposure, right, which in
04:30:35many
04:30:36cases
04:30:37determine more the risk than
04:30:39the other
04:30:40portions. So we have to
04:30:41deal with this very, very
04:30:43long,
04:30:45list of potential risk factors
04:30:47to try to estimate
04:30:49the probability of a certain
04:30:51trait.
04:30:52And we have to do
04:30:53that in,
04:30:55operating and compute systems that
04:30:57do not allow data to
04:30:59come out. So you have
04:31:00to do the opposite. You
04:31:02distribute the computation,
04:31:04tap into these resources, and
04:31:05and try to create,
04:31:08these equations
04:31:09or
04:31:10machine learning and so on.
04:31:12So this is a center
04:31:13that we we lead from
04:31:14Yale. It involves,
04:31:16also,
04:31:18colleagues at the Broad Institute
04:31:20and at,
04:31:21UC San Diego primarily.
04:31:24And what it does is
04:31:26try to tap into the
04:31:27All of Us Research Program,
04:31:29which is a large, large
04:31:30database
04:31:32of, consented participants
04:31:34that, have whole genome sequence
04:31:36data, more than four four
04:31:38hundred k of such data.
04:31:40And the million veteran program,
04:31:42which is an equivalent resource,
04:31:44which is also kind of
04:31:46sequestered from,
04:31:48the mainstream. It has to
04:31:50has its own compute environment,
04:31:52and, it's reaching the million
04:31:55mark very quickly. So how
04:31:56do we,
04:31:58compute with this data in
04:32:00a secure way, in a
04:32:02privacy preserving way, has been
04:32:04one of the,
04:32:05topics of our research.
04:32:08Now I'll go back to
04:32:10another thing that might be
04:32:11more exciting related to generative
04:32:15AI.
04:32:16So
04:32:17that classification problems I I
04:32:19told you, people criticize
04:32:21AI by saying, well, it's
04:32:22not very creative.
04:32:24It just recognize patterns that
04:32:25you give examples for. Right?
04:32:27So how can it be
04:32:28more creative? And you can
04:32:30see here, you might like
04:32:31it or dislike it. And
04:32:33you might see that there
04:32:34is a source
04:32:35from which it tries to
04:32:38be
04:32:39creative from. Right? So think
04:32:40about this image
04:32:42for a while.
04:32:43And also this one, this
04:32:45is an old NVIDIA,
04:32:48research project that showed that
04:32:50from sources, the real people
04:32:52on the top, you can
04:32:54create the fake people at
04:32:56the bottom, and they might
04:32:57become indistinguishable,
04:33:00from
04:33:01for us to decide who
04:33:03are the real people. You
04:33:04mix them all together, and
04:33:05you might not be able
04:33:06to,
04:33:08because the
04:33:09technology got so good that
04:33:11it it does produce,
04:33:14fake examples that look
04:33:16exactly like the real ones.
04:33:18So generative
04:33:19adversarial networks, there are two
04:33:21networks that are adversarial,
04:33:23and one wins.
04:33:25One loses
04:33:26on an iterative fashion.
04:33:28One produces the fake data.
04:33:31The other one tells, oh,
04:33:32this is real. This is
04:33:33not. So the whole game
04:33:35here is to generate and
04:33:37keep generating until
04:33:40the generator wins and the
04:33:41classifier can no longer say,
04:33:44you don't know. Now I'm
04:33:45fifty fifty.
04:33:46So I put here pardon
04:33:47my
04:33:48graphics because they're not professional,
04:33:50but they they give an
04:33:51idea here. You have a
04:33:53real observation. You have some
04:33:55noise, and the generator creates
04:33:57a fake observation.
04:33:59Right? That goes to the
04:34:01classifier
04:34:02network.
04:34:03The classifier network
04:34:05takes that takes a real
04:34:06observation, and it's supposed to
04:34:08say, oh, this is real.
04:34:09This is fake. So in
04:34:10this first example here,
04:34:13it does exactly that. So
04:34:14it wins, right, because it
04:34:16it still,
04:34:18can discern.
04:34:19This is another example. You
04:34:21keep doing it and says,
04:34:22oh, this is fake. This
04:34:23is real.
04:34:25Still wins
04:34:26until you get to this
04:34:27point where it says, well,
04:34:29this is these are both
04:34:31real.
04:34:32And and at that point,
04:34:33the generator wins. You stop
04:34:36training, and that's how you
04:34:37have your model.
04:34:38Right? So a large language
04:34:40model is, you know, with
04:34:41all the simplifications
04:34:43is
04:34:44nothing but the generator part.
04:34:46So what you have, you
04:34:48have what's called a prompt,
04:34:49which is an input. And
04:34:50you say diagnose a person
04:34:52with chest pain or pain
04:34:53smoker,
04:34:54and the large language model.
04:34:56So this is Chachipiti.
04:34:58We'll start. I'm not a
04:35:00doctor, but here's the things
04:35:01that you need to consider
04:35:02and so on. So it's
04:35:04just the generator now. You're
04:35:05not training anything anymore.
04:35:08And the generator, what it
04:35:09does in this large language
04:35:11models is guessing the next
04:35:13next word that would come
04:35:16given the prompt that you
04:35:17you gave. So it, ingested
04:35:20a whole lot of the
04:35:21information on the Internet,
04:35:23on articles, and so on,
04:35:25and now is able to
04:35:26generate something like this.
04:35:29So
04:35:31you take another example here,
04:35:32and it would come up
04:35:34with things that for the
04:35:35nonexpert seems pretty reasonable. It's
04:35:38it's not a total,
04:35:40out of line or hallucination,
04:35:43which the initial,
04:35:45large language models would hallucinate
04:35:47much more than now. They
04:35:49would create things, and they
04:35:50still do. But if you're
04:35:52not the expert, it looks
04:35:53pretty
04:35:54polished and and so on.
04:35:57So,
04:35:58what are health care
04:35:59organizations doing with generative AI?
04:36:02So that,
04:36:03application of ambient listening
04:36:05has some component
04:36:07of that because it generates
04:36:08the test text for you
04:36:10as well as the
04:36:12the one that does,
04:36:14draft replies for the inbox.
04:36:16But there's a whole lot
04:36:17of other things happening
04:36:20in terms of clinical research
04:36:22that, as you know, that
04:36:24article that said,
04:36:26large language models pass the
04:36:28exams and and so on,
04:36:30because it ingested a lot
04:36:32of those questions, right, in
04:36:33the textbooks and everything. And
04:36:35it's,
04:36:36it memorized them, and it's,
04:36:39sending back what it learned
04:36:41from it.
04:36:43So published medical cases. They
04:36:45use biomedical literature.
04:36:47Now the problem is they
04:36:48cannot use real observations
04:36:50or should not. Even though
04:36:52there
04:36:53are, you can purchase electronic
04:36:55health records,
04:36:56these days, so I I
04:36:57bet they used some of
04:36:58those too.
04:37:00But what we wouldn't do
04:37:02is have our records,
04:37:05be input into CheckGPT.
04:37:08And so and and no,
04:37:10organization
04:37:10does that because they have
04:37:12this ability to to memorize.
04:37:14Right? And if they have
04:37:15the ability to memorize,
04:37:17they can return the case
04:37:19exactly as it is, and
04:37:20this can be a privacy
04:37:22violation. So it's an important
04:37:24thing to note
04:37:25that
04:37:26the the problem is they
04:37:27don't learn anything.
04:37:29Right? We have,
04:37:31the the health care organizations
04:37:33have some contractual agreements that
04:37:35allow
04:37:36to be the data, to
04:37:38be used, but not to
04:37:39learn with the data, not
04:37:41to keep anything
04:37:42that the user is entering
04:37:45in that. And that's the
04:37:46the
04:37:47correct use to do through
04:37:49the organization.
04:37:50But on the other hand,
04:37:52it it won't learn. Right?
04:37:53So there is a problem
04:37:55right there. We we have
04:37:56some, developments in
04:37:59developing our own,
04:38:01medical large language model. We
04:38:03call that,
04:38:04medical lemma.
04:38:06And the reason is lemma
04:38:08is a foundational model. It's
04:38:10one such large language model
04:38:12from
04:38:13from Meta, from a former
04:38:15Facebook,
04:38:16in which they they made
04:38:17it available, opens,
04:38:20open to the more training
04:38:22and so on. So you
04:38:23can do that, but it's
04:38:24still, you know,
04:38:25very initial and still we're
04:38:27not doing
04:38:28with,
04:38:29the records because of that,
04:38:31privacy concerns that I mentioned
04:38:34before.
04:38:34So the challenges here at
04:38:36privacy are the black box
04:38:38nature of AI and and
04:38:40this myth that it will
04:38:41take over the world and
04:38:42and,
04:38:43you know, rule it by
04:38:44itself.
04:38:45Computational
04:38:46infrastructure
04:38:47is very demanding, so so
04:38:48we cannot do in academic
04:38:50medical centers the same way
04:38:52as industry
04:38:53can.
04:38:54Workforce is is better paid
04:38:56in nonacademic
04:38:58environment. So those all those
04:38:59AI, you know, super experts
04:39:01are working in industry and
04:39:03not in the in academia
04:39:05right now. And that there's
04:39:07the whole issue of AI
04:39:08fairness and unfairness,
04:39:10some of which is the
04:39:12same unfairness that any logistic
04:39:14regression model would have, but
04:39:16no one talked about that.
04:39:18And now everyone talks about
04:39:20as if AI had this,
04:39:23way more unfair.
04:39:24It it is unfair. I
04:39:26mean, I I can say
04:39:27it's as unfair as the
04:39:29sources of data it it
04:39:31got.
04:39:31And you can play tricks
04:39:33with it, and you will
04:39:34realize
04:39:35it can give different answers
04:39:37depending on
04:39:39how it learned that certain
04:39:40things are,
04:39:42you know, are more,
04:39:44there is more written about
04:39:47a a false
04:39:48claim than a a real
04:39:49claim. So it it might
04:39:51come up with this,
04:39:53issues as well. So here,
04:39:55I would say,
04:39:57unfairness comes from several
04:39:59places in including the misuse
04:40:01of data. This is a
04:40:03very old example, but it's
04:40:04still,
04:40:05coming back,
04:40:07whenever we talk about,
04:40:09why American Indian, American native
04:40:11populations are not so interested
04:40:14in participating in research
04:40:16because,
04:40:17it can be portrayed very
04:40:18negatively.
04:40:19Right?
04:40:20The other one was the
04:40:22the Havasupai
04:40:23tribe in Arizona. As you
04:40:25know, they, authorized
04:40:27for,
04:40:28use of data for a
04:40:30diabetes study, and then suddenly,
04:40:32there were all other studies,
04:40:34mental health, and and other
04:40:36things
04:40:37done without data. So so
04:40:38it was a,
04:40:40a violation.
04:40:41And, also, the limited benefit,
04:40:43if the populations don't see
04:40:45somewhat a benefit
04:40:47that,
04:40:48applies to their particular,
04:40:51group, they're less likely to
04:40:53want to participate.
04:40:55So AI created a lot
04:40:56of this,
04:40:58angst and fear because the
04:40:59first self,
04:41:01driving vehicles,
04:41:03they were not trained in
04:41:04different skin tones. They were
04:41:05not trained on height,
04:41:08of people,
04:41:09and they could make mistakes,
04:41:12you know, around those. So
04:41:14the
04:41:16this,
04:41:18headline came up. Another one
04:41:21is about this one in
04:41:22which,
04:41:23someone wanted to optimize
04:41:25medical appointments
04:41:27and and use the variety
04:41:29of variables,
04:41:30but it ended up,
04:41:32overbooking
04:41:34patients who were from a
04:41:36certain,
04:41:37racial and ethnic
04:41:38background
04:41:39because
04:41:40it predicted that there will
04:41:42be more no shows,
04:41:44according to that. So the
04:41:45equation would predict that. And
04:41:47you can see how the
04:41:49problematic
04:41:49that that can become. So,
04:41:51again, it's not AI's fault.
04:41:53It's not
04:41:54the regression fault. It's the
04:41:55fault of whoever is employing
04:41:58this and how they utilize
04:42:00it. And there are several,
04:42:02studies on AI chatbots. The
04:42:04things that it says back
04:42:06to us, particularly the older
04:42:08versions,
04:42:09can be very disturbing
04:42:11because it learned from the
04:42:13Internet.
04:42:14Right? So it's it's a
04:42:15problem. It's a real, real
04:42:17problem.
04:42:19Now we tried to,
04:42:21gather a a small group
04:42:22of,
04:42:23industry and academic
04:42:25medical center leaders to talk
04:42:27about this
04:42:28upfront
04:42:29and to talk about,
04:42:31potential
04:42:33ways to address the impact
04:42:35of this or or redress
04:42:37the populations that were affected
04:42:39by certain kinds of algorithms.
04:42:42We didn't say
04:42:43AI because AI is in
04:42:44the midst of many others
04:42:46that actually are much more
04:42:47used
04:42:48these days.
04:42:50So I also wanted so
04:42:52I talked about the challenges
04:42:53and the opportunities. There are
04:42:54many, many things.
04:42:56Again, talking about precision medicine,
04:42:58about federated
04:43:00learning. So as, Roy mentioned,
04:43:04I I did,
04:43:06manage to get the University
04:43:08of California system to collaborate
04:43:10on electronic health records,
04:43:12which, wasn't the case,
04:43:14before.
04:43:15And,
04:43:17the the reason is that
04:43:18we did it in such
04:43:19a way that was fair
04:43:21to to everyone and it
04:43:23protected the privacy as well.
04:43:25So it was a teamwork,
04:43:27with
04:43:28a joint governance model and
04:43:30open budget.
04:43:31Do not,
04:43:33underestimate
04:43:33the importance of this open
04:43:35budget
04:43:36portion of it because it
04:43:37was a deterrent
04:43:39to collaborate.
04:43:41Like, oh, this institution is
04:43:42getting more money than the
04:43:44other one and so on.
04:43:45So that's a very important
04:43:46thing. You need to have
04:43:47a common data model with
04:43:49the decentralized analysis
04:43:51algorithms,
04:43:52ability to opt out of
04:43:54research studies, and then,
04:43:56the sites help each other
04:43:58to live lift all boats,
04:43:59but it takes a while
04:44:00for people to get in
04:44:02in this spirit.
04:44:04We later,
04:44:06extended this to the whole
04:44:07VA network of electronic health
04:44:09records
04:44:10as well as another network
04:44:12in Los Angeles so we
04:44:14could have a much larger,
04:44:16universe of people.
04:44:18And,
04:44:20again, we also wanted to
04:44:22go beyond
04:44:23being data users, which we
04:44:25definitely are, but also to
04:44:27be data producers. How can
04:44:29we understand the process of
04:44:30recruiting these people when getting
04:44:32the data
04:44:33and so on? So so
04:44:34we
04:44:36created
04:44:37a a consortium
04:44:38to respond to the all
04:44:40of us,
04:44:41program,
04:44:43request for applications
04:44:44in developing recruitment centers, and
04:44:47that's how we we did
04:44:48it there. These are the
04:44:49main
04:44:50centers
04:44:51involved in our case.
04:44:53And we were one of
04:44:55several consortia that,
04:44:57were created across the country
04:45:00to do that,
04:45:02including,
04:45:02federal
04:45:03federally qualified health care centers,
04:45:05which are more community oriented,
04:45:08regional medical centers and and
04:45:10other academic
04:45:11medical
04:45:12centers as well.
04:45:14So when, we did all
04:45:15that and then COVID hit
04:45:17and all of us programs
04:45:18stopped because no one could
04:45:20be recruited.
04:45:21However, we could still use
04:45:23the the the network that
04:45:25we had to do other
04:45:26kinds of work. So we,
04:45:29decided that if there were
04:45:31clinical questions that were answerable
04:45:33by electronic health records, we
04:45:34would try to pull our
04:45:36systems to do. So we
04:45:38got our partners plus some
04:45:39others who wanted to participate,
04:45:42including a a large, system
04:45:44in Germany.
04:45:45And and in twenty twenty,
04:45:47we were,
04:45:49asking questions of this data
04:45:51of some,
04:45:53again, from an initial
04:45:55type of phase. One question
04:45:56that came early on was,
04:45:58should we stop ACE inhibitors
04:46:01for hypertensive
04:46:02patients because of potential,
04:46:05issues with the virus,
04:46:08receptors and so on. And
04:46:10we we were able to
04:46:11show that with our data
04:46:13for hospitalized patients,
04:46:15it seemed that ACE inhibitors
04:46:17were doing just fine, where
04:46:19the mortality
04:46:20of those patients was lower
04:46:22than ARBs and, other medications,
04:46:25and it was definitely
04:46:26all of them better than
04:46:28stopping any medication in this
04:46:30case. So in this case,
04:46:32there were ten institutions
04:46:33involved, and, we were able
04:46:35to,
04:46:36do this early
04:46:39finding that later got confirmed
04:46:41in observational
04:46:42studies. So that that was
04:46:44very reassuring.
04:46:46We did, you know, varied
04:46:47analysis of this sort in
04:46:49which,
04:46:51in this case,
04:46:53you could,
04:46:54say the mortality,
04:46:55for example,
04:46:57of,
04:46:59according to Brace,
04:47:00was not different for hospitalized
04:47:02patients in our system,
04:47:05whereas in general, there was
04:47:07this description.
04:47:08And the reason is that
04:47:10once you adjust for age,
04:47:12the other factors become much
04:47:14less important. So we were
04:47:16able to do this multivariate,
04:47:19analysis
04:47:20again with data that are
04:47:21all distributed. We never got
04:47:23the data centralized, and we
04:47:24were able to answer very,
04:47:26very simple things adjusting for
04:47:29confounders
04:47:30and so on.
04:47:32So
04:47:34since I moved, we we
04:47:35started also,
04:47:37Yale as one of those
04:47:39recruitment centers and also our
04:47:41partner in Puerto Rico.
04:47:43We had to name change
04:47:44the name of the consortium,
04:47:46the Coast to Coast Consortium,
04:47:48but we continue to do
04:47:49the same thing. And and
04:47:51it's a large program. It's
04:47:52very exciting
04:47:53to be in it.
04:47:55So I talked about federated
04:47:57learning on the a side
04:47:59is how different hospitals have
04:48:02data
04:48:03and can collaborate.
04:48:04There's also
04:48:06what's called vertical partitioning of
04:48:08the data. It's the hospital
04:48:09has some data from the
04:48:10patient.
04:48:11The wearable company has some
04:48:13other data. The, genome sequencing
04:48:15company has another one. So
04:48:17it's what's called vertical partitioning,
04:48:19and we can do
04:48:21federated learning, federated analysis in
04:48:23this
04:48:24similar way. It's harder. Actually,
04:48:26this case is much harder
04:48:28than the first one.
04:48:30And then we have,
04:48:32work on,
04:48:33privacy protection,
04:48:34on data standardization,
04:48:36and also in informed consent.
04:48:39And the people
04:48:40want to contribute. Do they
04:48:41wanna contribute everything or they
04:48:43want to omit certain
04:48:45areas or they want to
04:48:47contribute only if this type
04:48:48of research, but not that
04:48:50type of research or this
04:48:51type of research entity
04:48:54and not another one. So
04:48:56all sorts of variations
04:48:58there, and we use a
04:49:00blockchain
04:49:01technology
04:49:01to keep track of all
04:49:03these preferences.
04:49:05The vertically partitioned so that
04:49:07some, data, for example, in
04:49:09Texas,
04:49:11whereas the phenotypes, the electronic
04:49:13health records were in Oklahoma,
04:49:15so we developed the software
04:49:16so that they could do
04:49:18do the analysis without putting
04:49:21the two together.
04:49:22So the our vision essentially
04:49:24is to leave no one
04:49:25out,
04:49:26and have this large consortia
04:49:28to increase,
04:49:30the representation
04:49:31of underrepresented
04:49:33groups. The All of Us,
04:49:34in particular,
04:49:35is very,
04:49:38very focused on having groups
04:49:40that were previously underrepresented.
04:49:43And then in our case,
04:49:45the the the technology people,
04:49:47we want to develop new
04:49:49methods and tools that allow
04:49:51this research findings to be
04:49:52applicable
04:49:54to all. So
04:49:55the the consortia
04:49:57help link across disciplines,
04:49:59and, also, we do training,
04:50:02which I will talk about
04:50:03in,
04:50:04in a moment.
04:50:06We also participate,
04:50:08as the coordinating center for
04:50:10this
04:50:11relatively small program from NIH.
04:50:13It's called Bridge to AI
04:50:15in which the goal is
04:50:16to produce data that can
04:50:18be used,
04:50:20by machine learning researchers, by
04:50:22AI researchers
04:50:24in a way that advances,
04:50:26the field. So data sharing
04:50:28and access, we are dealing
04:50:29with a lot of,
04:50:31considerations on the ethical and
04:50:33legal implications, the provenance of
04:50:35data, technical needs,
04:50:38basic data sharing.
04:50:40So I would say what
04:50:41we're trying to do,
04:50:43is,
04:50:44create a master's program because
04:50:45we are also short of
04:50:46personnel
04:50:47that can do all of
04:50:48this. So we're proposing
04:50:50one, which is also in
04:50:52response to the,
04:50:54Yale task force
04:50:55on AI that called for
04:50:57new programs,
04:50:58called for new,
04:51:01training,
04:51:02in particular,
04:51:03and, is delivering on that
04:51:06by having also remember I
04:51:08said the hardware
04:51:09is only in industry and
04:51:11so on. So we will
04:51:12get some of that, and
04:51:13we will be able to
04:51:14do some,
04:51:15smaller scale but very impactful
04:51:18research
04:51:19in that area.
04:51:21Benefits to Yale, we are
04:51:22we're doing a certificate
04:51:24program already.
04:51:25I wanna move on to
04:51:26this master's program
04:51:28and train with another departments
04:51:30and and then work with
04:51:32Yale Ventures and and everything.
04:51:34So benefits to New Haven,
04:51:36we we think we can
04:51:37extend these courses. Many of
04:51:39them are online courses,
04:51:42to other institutions,
04:51:43regionally or nationwide.
04:51:45That's another college that approached
04:51:48us.
04:51:49They have a new school
04:51:50of applied computational sciences, so
04:51:52we're trying to,
04:51:54develop a program that,
04:51:57interacts
04:51:58with theirs. So here's my
04:52:00my last one. It's complicated,
04:52:02but we we can do
04:52:03it. There is a lot
04:52:05a lot of areas to
04:52:06to tap at the same
04:52:07time from,
04:52:09mistrust in AI to mistrust
04:52:12in research in general, biomedical
04:52:14research,
04:52:15to, yeah, we got a
04:52:17good model. Can we apply?
04:52:18Can we not apply?
04:52:20Health care organizations,
04:52:22trade offs between,
04:52:24you know, more openness
04:52:26of,
04:52:27of data to privacy concerns
04:52:30and also,
04:52:31competitiveness
04:52:32with other
04:52:33institutions. So a whole lot
04:52:35at play.
04:52:36So I would like to
04:52:37once again thank the the
04:52:38sponsors in this case, and
04:52:40thanks the organizer for the
04:52:42opportunity to talk to you.
04:52:53Time for a couple of
04:52:54questions.
04:52:55I'll start. That was fantastic,
04:52:57and I know we're doing
04:52:58a lot of work at
04:52:59Yale. When you came here,
04:53:00did you have to increase
04:53:00the computing capacity of the
04:53:02institution to to do all
04:53:03this, or was it already
04:53:04existing?
04:53:05No. We it is actually
04:53:07it,
04:53:07it has been planned and,
04:53:10purchased, and it will be
04:53:12implemented by end of this
04:53:14year. So it took took
04:53:15a little while, but it
04:53:16it's, it's going to be
04:53:17there. Because I imagine if
04:53:19we're using this for our
04:53:19notes, it has to be
04:53:20an internal system because we
04:53:22can't we can't use anything
04:53:23from the outside. Is that
04:53:24Exactly. So it's exactly that
04:53:25portion that I found at
04:53:27Yale could could benefit from
04:53:29some updates
04:53:31upgrades. Any other questions or
04:53:32comment? Nikolai, any any thoughts?
04:53:39That was that was really
04:53:40great.
04:53:41Can you comment on the
04:53:42classifier
04:53:43in the generated model? Are
04:53:44those
04:53:45algorithm based? Are they influenced
04:53:47by humans? Or
04:53:49that that still is a
04:53:50bit of a black box,
04:53:51to me. Can you Well,
04:53:52originally, they are just machine
04:53:54learning
04:53:55model classifiers. Right? But
04:53:58things have,
04:54:00changed in in many ways,
04:54:01so they're more complicated than
04:54:03what I,
04:54:05described. But but, essentially, the
04:54:07the classifier
04:54:08is a is a network
04:54:11is a neural network
04:54:13trying to essentially say, I
04:54:15I can no longer,
04:54:18distinguish. Right? But, of course,
04:54:20then humans come and say,
04:54:22yeah. You're right. Or maybe
04:54:24the classifier is not so
04:54:26so good. So I think
04:54:27that, like, the faces, right,
04:54:28in the beginning, they start
04:54:30not
04:54:31not doing the the right
04:54:33thing. It something is wrong.
04:54:36But over time because the
04:54:37classifiers have the gold standard.
04:54:40Right? So so to have
04:54:41the gold standard, you had
04:54:43to have some people
04:54:44to,
04:54:46to judge that. There are
04:54:47also some other methods in
04:54:49which that that you need
04:54:52fewer gold standards in order
04:54:54to do that. But, essentially,
04:54:56at some point, you have
04:54:58to agree that it did
04:54:59the right thing.
04:55:00Right.
04:55:01Any more questions?
04:55:03I'll ask a question.
04:55:04So,
04:55:05the AI has been,
04:55:08utilized in, imaging analysis as
04:55:10well as in, you know,
04:55:12not only in radiology, but
04:55:13also in pathology.
04:55:14Mhmm. And I wonder, you
04:55:16know, so how does it
04:55:17affect the the workforce issues?
04:55:19You know, does it help,
04:55:20to
04:55:23relieve some of the burden
04:55:24because we have more and
04:55:25more images to analyze, more
04:55:27and more pathology reports to
04:55:28generate. So I just don't
04:55:29know what's the status of
04:55:31the application of this type
04:55:32of AI in our everyday
04:55:34work because it is used.
04:55:35Right? So I know that
04:55:36images are red, and it's
04:55:38interesting how physicians who are
04:55:40responsible for it interact with
04:55:41it and, you know, so
04:55:42how much verification
04:55:44is going on. Do they
04:55:45have to verify everything?
04:55:47I would say,
04:55:49not everything, but it's advisable
04:55:52that, you know, you agree
04:55:54with with what's in there.
04:55:56Focus in your work. You
04:55:57know? Look more at these
04:55:58images. Look more at this
04:56:00area because that's where there
04:56:02could be,
04:56:04some some findings that you
04:56:05you have not reported
04:56:07yet.
04:56:09In other
04:56:11not trivial, but, important things,
04:56:13segmentation of the area. And
04:56:15is is it growing over
04:56:16time? Is it not growing
04:56:17over time? So we tell
04:56:18you, you have to have
04:56:20the the particular
04:56:22nodule you're worried about, and
04:56:25the machine would segment it
04:56:27for you. The machine would
04:56:28get the other image and
04:56:30register to say that's this
04:56:32is the nodule that is
04:56:33growing, and it's growing at
04:56:35this rate. And so so
04:56:36I think it it can
04:56:37assist you in some
04:56:39aspects of it,
04:56:41and that's the exact goal.
04:56:43Right? Is the,
04:56:46on the imaging side, even
04:56:47if you only have the
04:56:48the imaging,
04:56:50it it helps you,
04:56:53focus on certain areas.
04:56:56Fantastic. Any more questions?
04:56:58This was really a wonderful,
04:57:00talk. I think it leaves
04:57:01a lot of room for
04:57:02thought, and there are many
04:57:03people here who I'm sure
04:57:05will be contacting you now
04:57:06for, working with clinical applications.
04:57:08Thank you so much. Thank
04:57:09you.
04:57:15So now, I'm gonna invite
04:57:16Scott Huntington,
04:57:18our chief quality officer.
04:57:20So expect a quality session
04:57:21there, my friend. And he's
04:57:23gonna lead us through,
04:57:24and one of our organizers
04:57:26is here. He's gonna lead
04:57:26us through, the section on.
04:57:38Thanks, Roy and the organizers.
04:57:41So it's it's really my
04:57:42pleasure to introduce my, great
04:57:44colleagues,
04:57:45with malignant hematology. We have
04:57:46three great speakers today.
04:57:48First up will be doctor,
04:57:50Sethi, who's currently covering our
04:57:51inpatient and was able to
04:57:53come from the hospital,
04:57:55to present the key updates
04:57:56on lymphoma.
04:57:57Doctor Sethi joined us from
04:57:58Vanderbilt a few years ago,
04:57:59focuses on aggressive lymphomas,
04:58:02and we'll have a great
04:58:03discussion.
04:58:04Next up will be doctor
04:58:05Browning,
04:58:06who will, present
04:58:07on multiple myeloma.
04:58:09She was trained here at
04:58:10Yale and also had a
04:58:11advanced fellowship in, amyloidosis
04:58:14before coming to Yale, and
04:58:15she'll have a great talk
04:58:17on myeloma.
04:58:18And then last will be
04:58:19doctor Podosev who will, talk
04:58:21about leukemia.
04:58:22We'll then move on to
04:58:23the, case discussion. We'll have
04:58:25three separate cases, and we'll
04:58:27invite doctor Gouda and doctor
04:58:28Souffy to join us, for
04:58:30their perspective on transplant CAR
04:58:32T for those patients. So,
04:58:33without any delay, doctor Sathy.
04:58:42Thanks, Scott.
04:58:44Thank you, Roy, and all
04:58:46the organizers and the sponsors.
04:58:48It is my pleasure to
04:58:49present the lymphoma heart topics
04:58:51in twenty twenty four on
04:58:52behalf of my,
04:58:54great lymphoma colleagues.
04:58:56So,
04:58:58these are my disclosures.
04:59:01So fortunately, we have a,
04:59:03you know, ton of new
04:59:04data, a lot of exciting
04:59:06updates in lymphoma. In the
04:59:07interest of time, I'm going
04:59:08to focus on a few
04:59:10in diffuse large B cell
04:59:11lymphoma and classical Hodgkin lymphoma.
04:59:14So we will start with
04:59:15diffuse large B cell lymphoma
04:59:17looking at
04:59:19four
04:59:20studies. Three of them are
04:59:22conference abstracts, and one of
04:59:24them actually was a published
04:59:25paper that I wanted to
04:59:26highlight.
04:59:28Alright. Starting with the first
04:59:29study, this is, EHA s
04:59:31two three nine, first data
04:59:33from subcutaneous epcaritumab
04:59:35and polatuzumab,
04:59:36with redox with polar r
04:59:39chip. Basically, epcuritumab
04:59:40and polar r chip for
04:59:42first line treatment of diffuse
04:59:43large b cell lymphoma.
04:59:45This is one of the
04:59:45arms from the EPCCORE
04:59:47NHL five study. So as
04:59:50you can see here, there
04:59:50are a lot of mAbs,
04:59:53in that title. So just
04:59:55wanted to highlight that,
04:59:57the epkiritumab
04:59:58here is it's a BiTE
05:00:00antibody that has two the
05:00:02two arms. One of them
05:00:03is directed against c d
05:00:04twenty, which, attaches to our
05:00:06lymphoma
05:00:07cells and then the c
05:00:08d three,
05:00:09two t cells.
05:00:11Polituzumab vedotin is, an
05:00:14antibody drug conjugate against seven
05:00:17c d seventy nine b.
05:00:18And then we all know
05:00:19rituximab
05:00:21are standard c d the
05:00:22most commonly used CD twenty
05:00:24antibody. Okay. The key eligibility
05:00:26criteria was that it you,
05:00:28included patients with untreated
05:00:30diffuse large B cell lymphoma
05:00:32and also included high grade
05:00:34B cell lymphoma, including those
05:00:35with high risk trans translocations
05:00:38as well as follicular lymphoma
05:00:40grade three b,
05:00:41and it included patients with
05:00:42IPI two to five.
05:00:45Here you would see that
05:00:47the inclusion criteria very similar
05:00:49to what was,
05:00:50the population in the polar
05:00:52r CHIP study, which, you
05:00:53know, has become standard of
05:00:55care for us, frontline for
05:00:57many of our DLBCL patients
05:00:59with high IPI intermediate to
05:01:01high IPI score. So as
05:01:03we see here, these patients
05:01:05had, IPI scores from,
05:01:07two to five.
05:01:08Very, few patients with those
05:01:10high risk,
05:01:11mutations, this
05:01:13but it did have a
05:01:14balance of both GCB and
05:01:16non GCB subtype of diffuse
05:01:17large B cell lymphoma.
05:01:20So
05:01:21looking at the so, again,
05:01:23this is very early data.
05:01:24This is, at a median
05:01:25follow-up of seven point four
05:01:27months.
05:01:28They report,
05:01:29that the, overall response rate
05:01:31was hundred percent and the
05:01:32CR rate was eighty eight
05:01:33point six percent.
05:01:36So median time to response
05:01:37was two point seven months,
05:01:40and, the median time to
05:01:42CR was two point eight
05:01:43months. As you as you
05:01:44can see, most people,
05:01:45did end up getting a
05:01:47a CR. And the, at
05:01:49a again, at a median
05:01:50follow-up of seven point four
05:01:51months, the,
05:01:53median duration of CR is
05:01:55not reached.
05:01:57So,
05:01:58as so there are only
05:01:59thirty five patients. The study
05:02:00is really not powered to
05:02:02look at subgroups.
05:02:03But,
05:02:05so, we didn't really it
05:02:06didn't really show any differences
05:02:08in the studied groups shown
05:02:10here.
05:02:12And then also,
05:02:13in terms of
05:02:15toxicity,
05:02:16it were the most common,
05:02:20treatment,
05:02:20evident,
05:02:22emergent adverse events
05:02:24were basically cytopenias,
05:02:27and then also diarrhea
05:02:29and nausea, but that was
05:02:31a low grade. The most
05:02:32common, you know, higher grade
05:02:34were basically a pan
05:02:36different cytopenias.
05:02:38Importantly, there was no ICANS,
05:02:40and, CRS
05:02:42was low grade and predictable.
05:02:44This is what we see
05:02:44in most BiTE antibodies. We,
05:02:47the CRS is more predictable
05:02:48as compared to CAR T
05:02:49cell therapy,
05:02:51which is expected, and, most
05:02:53CRS events were,
05:02:55confined to cycle one.
05:02:57And,
05:02:58again, there was one death,
05:02:59and that was due to
05:03:01a urosepsis that was thought
05:03:02to be unrelated to epkoritumab.
05:03:05So, again, this is just
05:03:06one.
05:03:07This is a single arm
05:03:08study data. I think what
05:03:09we get from here that
05:03:11it is, possible to build
05:03:12on
05:03:13polar archipelagic.
05:03:15We have a safety signal
05:03:16for combining BiTE,
05:03:18with polar r chip from
05:03:19the study.
05:03:20But, really, we need to
05:03:21look at,
05:03:23more randomized data, and we
05:03:24do have, just this week,
05:03:26we have a trial with
05:03:27glofitumab, which is another,
05:03:29c d twenty, c d
05:03:30three BiTE antibody,
05:03:32and that is being comp
05:03:33glofitimab plus polar chip, it's
05:03:35being compared with polar chip
05:03:37in a randomized,
05:03:39study, and, we have that
05:03:40open at Yale now. So
05:03:41we're looking forward to enrolling
05:03:43patients on that.
05:03:46Okay. So,
05:03:48keeping with the BiTE theme,
05:03:50I now and glofetimab.
05:03:52So I'm going to talk
05:03:53about this, study next. This
05:03:55was also an EHA late
05:03:56breaking abstract. It was the
05:03:58STARGO trial. I'm
05:04:01sorry.
05:04:02Glofetimab plus gemcitabine,
05:04:04oxaliplatin, that is GLOFID gem
05:04:06mox compared with,
05:04:09and that was,
05:04:10for
05:04:11relapsedrefractory
05:04:13diffuse large b cell lymphoma.
05:04:14And this was being compared
05:04:15with our Gemox.
05:04:17So as you can see
05:04:17here, Glofed is also a
05:04:19c d three,
05:04:21c d twenty bite antibody,
05:04:23but this one has a
05:04:24specific two is to one
05:04:25format so it's, the inclusion
05:04:27key inclusion criteria were any,
05:04:39they just required one prior
05:04:41line of therapy.
05:04:43And,
05:04:45those that had were, had
05:04:47had had only one prior
05:04:49line of therapy had to
05:04:50be transplant and eligible.
05:04:52And,
05:04:53they stratified based on whether
05:04:55it was relapsed disease or
05:04:57refractory disease and then number
05:04:58of lines of therapy.
05:05:00And as I mentioned, this
05:05:01is a randomized study.
05:05:04We