Pathology Grand Rounds Video March 5, 2026 - Jason L. Hornick, MD, PhD

Transcript

• 00:00Good afternoon, everyone. Thank you
• 00:02so much for coming.
• 00:04Today is, our annual William
• 00:07McAllister Junior,
• 00:09Memorial Lecture.
• 00:11And,
• 00:12we are very
• 00:13honored to have, doctor Jason
• 00:15Hornick as,
• 00:16a recipient of this lectureship
• 00:18award.
• 00:19Doctor Hornick is a professor
• 00:21of pathology at Harvard Medical
• 00:23School, and our cancer chair
• 00:24at Brigham at Women's Hospital,
• 00:27director of anatomic pathology and
• 00:28immunohistochemistry
• 00:29at Brigham and Women's Hospital,
• 00:31and chief of soft tissue
• 00:32and bone pathology at, Mass
• 00:34General Brigham. So it's a
• 00:36great honor,
• 00:37to have doctor Hornick with
• 00:38us today. And,
• 00:40just a few words about
• 00:41this lectureship,
• 00:42and,
• 00:43I have to tell to
• 00:45to Jason, actually, our faculty
• 00:47voted for you to receive
• 00:48this lectureship award. It was
• 00:50a nomination process, and it
• 00:51was a vote. So it's
• 00:53really, you know, we wanna
• 00:54express how much we actually
• 00:55appreciate what you did for
• 00:57the field of surgical pathology.
• 00:59But back to doctor McAllister,
• 01:01who was
• 01:02a graduate of Yale College
• 01:04and John Hopkins Medical School
• 01:06and Department of Pathology. He
• 01:08actually came back to New
• 01:09Haven
• 01:10and served as the chief
• 01:11of surgical pathology for the
• 01:12new year,
• 01:13for the New Haven, hospital
• 01:15for twenty five years
• 01:17from, nineteen fifty three to
• 01:19nineteen seventy eight.
• 01:21And,
• 01:22he was diagnostic pathologist
• 01:25educator,
• 01:26and,
• 01:27his teaching including the science
• 01:28of medicine, also the art
• 01:30of living. He was an
• 01:31extraordinary
• 01:32mentor, adviser,
• 01:34friend to students, residents, coworkers,
• 01:36and physicians from all discipline.
• 01:38And I think this also
• 01:39described doctor Jason Hornick. So,
• 01:41I don't think doctor Jason
• 01:42Hornick needs any introduction, but
• 01:44just going to say a
• 01:45few words about you.
• 01:47Doctor Horney has started,
• 01:49his,
• 01:50medical school and completed his
• 01:52medical training at University of
• 01:54California, where she obtained his
• 01:55MD and the PhD degree.
• 01:57From West Coast to he
• 01:59moved to East Coast to,
• 02:00Harvard Medical School, where he
• 02:03did a training in anatomic
• 02:04pathology
• 02:05and,
• 02:06impressive fellowships in GYN, anatomic
• 02:09pathology, soft tissue pathology, hematopathology,
• 02:12and the gastrointestinal pathology.
• 02:13I don't see, I don't
• 02:14think we see that anymore.
• 02:16So, he's, like, one of
• 02:17the last, I would say,
• 02:19pathologist who actually went through
• 02:21such extensive training and is
• 02:22actually able to do
• 02:24ex excellent job in all
• 02:25of the areas.
• 02:27But what he is most
• 02:28known for is certainly the
• 02:29area of bone and soft
• 02:30tissue pathology and GI as
• 02:32well, but bone and soft
• 02:33tissue in particular,
• 02:34where he really defined multiple
• 02:36entities and actually help us
• 02:38to develop many diagnostic tests
• 02:40that we routinely use. So
• 02:41it's all thanks to you,
• 02:42Jason. So thank you so
• 02:43much.
• 02:44Numerous publications needed to say
• 02:46were five hundred. It's just
• 02:48I I stopped counting, actually,
• 02:50in your CV.
• 02:51But, other notable,
• 02:53contributions,
• 02:54he's been, editor of the
• 02:56WHO for the last three
• 02:57editions,
• 02:59and, he's currently serving as
• 03:01the president of the United
• 03:02States and the Canadian Academy
• 03:04of Pathology
• 03:05and is immediate past president
• 03:07of the Arthur Purdes South
• 03:08Society of Surgical Pathologists. So,
• 03:10Jason, thank you so much
• 03:12for, coming to us today
• 03:13and accepting this Lectureship Award.
• 03:15Thank you.
• 03:20Thank you so much, Sonia.
• 03:22And it's really such a
• 03:23pleasure to be here back
• 03:24at Yale.
• 03:25And I very much appreciate
• 03:26the faculty voting for me
• 03:27to come.
• 03:28I feel like I come
• 03:29about every ten years. It's
• 03:31been a while, but it's
• 03:32so nice to be back.
• 03:34I thought I'd come south
• 03:35and the weather would be
• 03:36amazing. But surprisingly, it's very
• 03:38similar to Boston.
• 03:40So, you know, throughout my
• 03:42career and throughout my
• 03:44kind of pathology years,
• 03:46I've really had a big
• 03:47interest in diagnostic immunistic chemistry.
• 03:50In my PhD, I worked
• 03:52on antibody engineering
• 03:54where, I was working on
• 03:55various constructs for tumor targeting
• 03:58before we had drugs that
• 03:59were antibodies, before Rituxan and
• 04:01Herceptin,
• 04:03and others.
• 04:04And I knew nothing nothing
• 04:05about pathology, but I knew
• 04:06a lot about antibodies. So
• 04:08it was sort of easy
• 04:09to become
• 04:11interested in immunohistochemistry.
• 04:13And in this session, I'm
• 04:14really gonna focus on how
• 04:16immunohistochemistry
• 04:17has changed over the last
• 04:18twenty years.
• 04:20I'm
• 04:21I like to kind of
• 04:23pretend there's a separation between
• 04:24the twentieth century and twenty
• 04:26first century.
• 04:27It's not, of course, a
• 04:28nice line there. But for
• 04:29the first twenty five to
• 04:30thirty years, we use this
• 04:32technology
• 04:34to try to define a
• 04:35line of differentiation.
• 04:37What do the tumor cells
• 04:38most closely resemble? And this
• 04:40still is critically important for
• 04:42us in
• 04:43oncologic surgical pathology and hematopathology.
• 04:47But now we have so
• 04:48many examples
• 04:50of
• 04:51using antibodies to recognize
• 04:53the protein correlates of molecular
• 04:55genetic alterations.
• 04:57So this is really a
• 04:59whole new way of using
• 05:01this technology, and I know
• 05:02there's many people here who
• 05:03are expert in this field
• 05:05as well.
• 05:06But I'm really gonna talk
• 05:08about molecular genetics, but only
• 05:10as what the protein
• 05:12surrogates are, the protein consequences
• 05:15for immunoskemetry.
• 05:17And,
• 05:18you know, we still use
• 05:19all the time so many
• 05:21antibodies that are helpful to
• 05:24to sub classify
• 05:25tumors that we can't
• 05:27diagnose down the microscope.
• 05:29Undifferentiated
• 05:30malignant neoplasms
• 05:32or narrowing down various diagnostic
• 05:34categories. But I'm not gonna
• 05:36talk about this. I'm gonna
• 05:37talk about what some people
• 05:38have liked to call next
• 05:40generation immunohistochemistry.
• 05:42And I've listed six different
• 05:44topics here, and I'm just
• 05:46gonna give one or two
• 05:47examples in each of these
• 05:49areas
• 05:50by way of introduction to
• 05:51this field.
• 05:52Some of you will be
• 05:54familiar with some of these
• 05:55antibodies, others will be new.
• 05:57And all the antibodies I'm
• 05:59talking about today are antibodies
• 06:02we use routinely in our
• 06:03clinical practice in my department.
• 06:06And I'm not gonna read
• 06:07this list because I'm gonna
• 06:09give examples as we go
• 06:10through them. But you can
• 06:11see the really broad range
• 06:13of applications
• 06:14this has opened up for
• 06:15us. In some cases, these
• 06:17antibodies can really help narrow
• 06:19down a differential diagnosis.
• 06:21But in other cases,
• 06:22they are perfect surrogates for
• 06:24molecular genetics,
• 06:26allowing us to perform a
• 06:27single antibody test for immunohistochemistry
• 06:30instead of any form of
• 06:31molecular genetic testing. And this
• 06:33is really valuable both for
• 06:35the turnaround time, the expense,
• 06:37and for practices throughout the
• 06:39world that don't have access
• 06:41to expensive next generation sequencing
• 06:43and other technologies.
• 06:45So that's in part been
• 06:46one of my goals of
• 06:48working in this field is
• 06:49to make it easy
• 06:50for pathologists to diagnose rare
• 06:53cancers
• 06:54in a very straightforward and
• 06:55rapid fashion.
• 06:57So probably the oldest example
• 06:59I'm gonna talk about today
• 07:00is an example where we
• 07:02can look for abnormal protein
• 07:04localization,
• 07:05and that is for beta
• 07:07catenin.
• 07:08I'm sure you all know
• 07:09about beta catenin, which is
• 07:11the protein,
• 07:12that is encoded by CTN
• 07:14and B1 gene
• 07:15involved along with the APC
• 07:18protein
• 07:19in the Wnt signaling pathway.
• 07:22And mutations
• 07:23in CTN and B1
• 07:25or biallelic inactivation of the
• 07:28adenomatous polyposis coli gene,
• 07:31most often in patients who
• 07:32have FAP, familial adenomatous polyposis,
• 07:36results in an aberrant localization
• 07:38of beta catenin
• 07:40from the cell membrane
• 07:41inside the cell to the
• 07:43cytoplasm and the nucleus.
• 07:45This is very helpful
• 07:47to support the diagnosis
• 07:49of a select
• 07:50group of tumor types
• 07:52that have mutations
• 07:54in one of these genes.
• 07:55And I've listed them here
• 07:57on this
• 07:58slide. And I'm just gonna
• 08:00show you two examples.
• 08:01So the first, a very
• 08:03old example.
• 08:04In nineteen ninety nine, it
• 08:05was discovered
• 08:07that beta catenin mutations
• 08:09are the most common drivers
• 08:12of desmoid fibromatosis.
• 08:14You know, so called deep
• 08:15fibromatosis
• 08:16that's a locally aggressive soft
• 08:18tissue tumor that arises fairly
• 08:20commonly,
• 08:22extra abdominal sites in the
• 08:23abdominal cavity.
• 08:25And, Liz Montgomery and colleagues
• 08:27showed us,
• 08:29twenty years ago now that
• 08:30we could use immunohistochemistry
• 08:33to look for aberrant localization
• 08:35of this protein as a
• 08:36very good diagnostic surrogate.
• 08:39It's not perfect. It's only
• 08:41about eighty percent
• 08:43sensitive for desmoid fibromatosis.
• 08:46So but fortunately it's negative
• 08:48in many other tumors we
• 08:49would consider
• 08:50in the differential diagnosis.
• 08:53But because specificity and sensitivity
• 08:55are not perfect,
• 08:57it must be interpreted in
• 08:58the context of morphology
• 09:00and clinical
• 09:02details.
• 09:03And at the same time,
• 09:05the lack of this pattern
• 09:06certainly does not preclude diagnosis.
• 09:09But this is great when
• 09:10you're dealing with a small
• 09:11core biopsy
• 09:12and you're feeling a little
• 09:13bit uncomfortable to make a
• 09:15firm diagnosis of this tumor,
• 09:17which obviously has very huge
• 09:19implications
• 09:20for management of the patient.
• 09:22So this is desmoid fibromatosis.
• 09:25This tumor is composed of
• 09:26these bland,
• 09:28uniform
• 09:28myofibroblastic
• 09:30spindle cells with
• 09:31tapering nuclei and indistinct cytoplasm
• 09:35arranged in long sweeping fascicles
• 09:37within a often collagenous stroma.
• 09:40One of the clues to
• 09:41the diagnosis is the presence
• 09:43of these
• 09:44relatively thick walled muscularized
• 09:47vessels
• 09:48that have variable perivascular
• 09:50edema between the bundles of
• 09:52spindle cells.
• 09:54And this is beta catenin.
• 09:55We get again, aberrant nuclear
• 09:57localization
• 09:59as a very helpful finding
• 10:01to diagnose desmoid fibromatosis.
• 10:04One other example is a
• 10:06tumor type that arises in
• 10:07the, sino nasal region that's
• 10:10called glomangio pericytoma.
• 10:12We used to refer to
• 10:13it as sino nasal hemangioparicytoma.
• 10:17This is a lesion that
• 10:18presents as a polypoid mass,
• 10:20most often in the ethmoid
• 10:21sinus or the nasal cavity.
• 10:24It can recur locally, but
• 10:25it's entirely benign. So certainly
• 10:27you don't wanna mistake this
• 10:28tumor for a sarcoma.
• 10:30This is a true tumor
• 10:32of perivascular
• 10:33contractile cells or pericytes,
• 10:36and it often
• 10:37expresses actin filaments.
• 10:40This is how they look.
• 10:41They often have these thin
• 10:43walled branching blood vessels.
• 10:45As you see in the
• 10:45middle of the field,
• 10:47the tumor cells are these
• 10:48very uniform and evenly distributed
• 10:51oval cells
• 10:52with very fine chromatin.
• 10:56In twenty fifteen,
• 10:58two groups simultaneously
• 11:00discovered
• 11:01that sino nasal glomangioparicitoma
• 11:05harbors
• 11:06activating mutations in beta catenin
• 11:08and CTN and B1,
• 11:10just like desmoid fibromatosis.
• 11:12And all of a sudden
• 11:13it became very easy
• 11:15to confirm the diagnosis with
• 11:17a single antibody.
• 11:19And it's much more impressive
• 11:21than desmoid tumors
• 11:22where you'll have really intense
• 11:25nuclear and cytoplasmic
• 11:26staining
• 11:27and essentially one hundred percent
• 11:28of the tumors of this
• 11:29class.
• 11:30So this is the way
• 11:32we now diagnose
• 11:33this uncommon tumor type.
• 11:37So moving on to cert
• 11:39looking for protein loss
• 11:41due to inactivation of tumor
• 11:43suppressor genes.
• 11:44This is not a complete
• 11:45list. There are many examples.
• 11:47You know, TP53
• 11:49obviously is an example we've
• 11:50known about for decades.
• 11:52That's a little bit different
• 11:54because of stabilization of the
• 11:55protein. We often get what
• 11:57looks like overexpression,
• 11:59but sometimes when you have
• 12:00truncating mutations,
• 12:02you do get a complete
• 12:03loss of the protein.
• 12:05But here's examples where we
• 12:06do find consistent complete loss.
• 12:09And again, I'm just gonna
• 12:10give you two examples.
• 12:12The first,
• 12:13example is BAP1.
• 12:15This is a nuclear deubiquitinase
• 12:18that was found in twenty
• 12:20eleven
• 12:21to be inactivated
• 12:23in the vast majority
• 12:25of pleural diffuse mesotheliomas,
• 12:29both in the sporadic setting
• 12:31and in patients who have
• 12:32germline mutations
• 12:34in BAP1, which is a
• 12:35very uncommon
• 12:37cancer predisposition
• 12:38syndrome.
• 12:40And since then, several groups,
• 12:42one of the first was
• 12:43led by Andy Cherg from
• 12:45Vancouver in Canada,
• 12:47showed us how we can
• 12:48use this antibody
• 12:49as a very helpful marker
• 12:51to diagnose mesothelioma
• 12:53when we don't have very
• 12:55clear invasion
• 12:57of adipose tissue of the
• 12:58parietal pleura and other structures.
• 13:01And those of you who
• 13:01work in surgical pathology know
• 13:04that that's really what you
• 13:05need to diagnose malignancy
• 13:08in atypical mesothelial
• 13:09proliferations.
• 13:12The BAP1 tumor predisposition
• 13:14syndrome
• 13:16predisposes
• 13:16not only to the development
• 13:18of mesothelioma,
• 13:20but to a range of
• 13:21very unusual melanocytic tumors,
• 13:23as well as renal cell
• 13:25carcinoma.
• 13:26And the sporadic counterparts of
• 13:28all these tumors
• 13:29can also harbor mutations,
• 13:31that are somatic.
• 13:35So this was a case
• 13:36I had a couple of
• 13:37years ago, which is a
• 13:38really nice illustration of this,
• 13:40finding.
• 13:41This was a patient who
• 13:42was,
• 13:43a seventy five year old
• 13:44man who had
• 13:46a plural effusion. Going like
• 13:48two thousand two thousand two
• 13:49thousand. Now it's just like
• 13:51we kind of We have
• 13:53can can someone please mute?
• 13:54Yeah. Two thousand total over
• 13:56four years. So with Someone
• 13:58help me mute them? Maybe
• 13:59not.
• 14:00Oh, well.
• 14:03And when the surgeon went
• 14:05in to look into the
• 14:06pleural cavity,
• 14:08she saw nothing abnormal. So
• 14:10she took these arbitrary biopsies
• 14:12all over the parietal pleura,
• 14:14and this was the only
• 14:15biopsy that had abnormality. You
• 14:17can see we have these
• 14:19uniform mildly atypical epithelioid
• 14:22cells,
• 14:23whether they're barely trickling into
• 14:25the fat,
• 14:26which would be how we
• 14:27would confirm the diagnosis of
• 14:29mesothelioma.
• 14:30But if this is all
• 14:31you get, it's a little
• 14:32bit scary to diagnose this
• 14:35almost uniformly fatal tumor. So
• 14:37let's look at BAP1 immunoscedochemistry.
• 14:40This was such a nice
• 14:41example. You can see the
• 14:42normal mesothelium
• 14:43on the upper and lower
• 14:45parts of the image on
• 14:46the right. And then we
• 14:47have the neoplasia in the
• 14:49middle
• 14:49and you see this really
• 14:51beautiful transition
• 14:52between normal nuclear staining
• 14:55and tumor cells that have
• 14:56a complete loss of BAP1.
• 14:58So this is a beautiful
• 14:59surrogate
• 15:00for biallelic inactivation of the
• 15:02BAP1 gene.
• 15:04And, my colleague, Lynette Scholl,
• 15:06along with David Chaple, who's
• 15:08now,
• 15:09works on faculty at the
• 15:10University of Michigan,
• 15:12evaluated a panel of different,
• 15:15antibodies that recognize,
• 15:17protein products of tumor suppressors
• 15:20and found that the combination
• 15:22not only of BAP1, but
• 15:23also of MTAP
• 15:24and Merlin, the NF2 gene
• 15:27protein product
• 15:28together can be very helpful
• 15:30to diagnose
• 15:31mesothelioma.
• 15:32And this has become a
• 15:33routine part of our diagnostic
• 15:35practice.
• 15:37One other marker I'm gonna
• 15:39mention is a tumor suppressor
• 15:41that is
• 15:42inactivated
• 15:43in a very rare aggressive
• 15:45cancer
• 15:46that occurs predominantly in the
• 15:47ovary of adolescent girls and
• 15:50young women
• 15:51that goes by this very
• 15:52long name, small cell carcinoma
• 15:54of the ovary
• 15:55of hypercalcemic
• 15:56type.
• 15:57It's a it's a long
• 15:58descriptive name. We often call
• 16:00it SCUT because it's easier
• 16:02than saying all these words.
• 16:04And we call it this
• 16:05name because in about two
• 16:06thirds of cases,
• 16:08patients present with quite profound
• 16:10hypercalcemia.
• 16:12And even with early stage
• 16:14disease localized to the ovary,
• 16:16their survival is very poor
• 16:18for this aggressive cancer type.
• 16:21Many cases have focal areas
• 16:23with these dilated follicle like
• 16:25spaces.
• 16:26That's very characteristic of this
• 16:28aggressive tumor
• 16:29and we call it small
• 16:30cell because most examples
• 16:33are dominated by fairly uniform
• 16:35small round blue cells with
• 16:36a little bit of cytoplasm.
• 16:38This is not really a
• 16:40small cell carcinoma as you
• 16:41think of small cell carcinomas
• 16:43of any other site.
• 16:45In fact, some people have
• 16:46proposed this is probably better
• 16:48aligned with a sarcoma,
• 16:49but we still call it
• 16:50this long descriptive name.
• 16:52In about fifty percent of
• 16:54cases,
• 16:55there is a large cell
• 16:56component
• 16:57which often has rhabdoid cytoplasmic
• 17:00inclusions,
• 17:01which are these intermediate
• 17:03filament containing
• 17:04perinuclear
• 17:05inclusions that look all hyaline.
• 17:07And that's looks very similar
• 17:09to malignant rhabdoid tumors of
• 17:11the of infancy.
• 17:14And again, we have multiple
• 17:16groups that essentially at the
• 17:17same time published in the
• 17:19same issue
• 17:20of Nature Genetics discovered
• 17:22that inactivating
• 17:23mutations in SMARCA4,
• 17:25which encodes a protein that's
• 17:27also called BRG1,
• 17:29are the defining genomic feature
• 17:31of this aggressive cancer type.
• 17:33And otherwise, they have a
• 17:35very
• 17:35flat genome. It's really a
• 17:37this isolated event
• 17:39can lead to this incredibly
• 17:41aggressive cancer type. And in
• 17:43fact,
• 17:44in about
• 17:45fifty percent of patients,
• 17:47even without a family history,
• 17:49they will be found to
• 17:50have a germline mutation. So
• 17:51the penetrance is quite variable.
• 17:54But this has now become
• 17:55an easy way to diagnose
• 17:57this rare cancer.
• 17:58We no longer have to
• 17:59send it to an expert
• 18:01to help us decide that
• 18:02this is in fact one
• 18:03of these rare cancers.
• 18:07And just like we saw,
• 18:08in BAP1,
• 18:10we have a complete loss
• 18:11of the normal nuclear
• 18:13reactivities
• 18:14for SMARCA4 or BRG1,
• 18:16and this is found in
• 18:18almost a hundred percent
• 18:19of tumors of this class.
• 18:23One example of a really
• 18:24nice epigenetic
• 18:25alteration we can assess for
• 18:27by immunoskemetry,
• 18:28and I talked about this
• 18:29with the trainees this morning,
• 18:31is a malignant peripheral nerve
• 18:33sheath tumor.
• 18:34As many of you know,
• 18:35about fifty percent
• 18:37of MPNSTs
• 18:38arise in patients with NF
• 18:40one,
• 18:41forty percent are sporadic,
• 18:43and ten percent arise following
• 18:45therapeutic radiation therapy.
• 18:47If you have a malignant
• 18:49spindle cell neoplasm
• 18:51in a patient with NF1
• 18:54or arising from a large
• 18:55nerve
• 18:56or coming out of a
• 18:57benign neurofibroma,
• 18:58the diagnosis is pretty easy.
• 19:00But outside of those associations,
• 19:03this is a really difficult
• 19:04diagnosis,
• 19:06especially because
• 19:08our markers of nerve sheath
• 19:09differentiation,
• 19:11the Schwann cell markers S100
• 19:13and Sox10,
• 19:14are only positive in just
• 19:16under half of the cases,
• 19:18and they're focal and weak
• 19:19and unimpressive when they are
• 19:21positive.
• 19:21So it really relies
• 19:23on,
• 19:24excluding other tumors you might
• 19:26consider in the differential diagnosis
• 19:29and pay paying close attention
• 19:31to the often very distinctive
• 19:33histology.
• 19:35Many examples of malignant peripheral
• 19:37nerve sheath tumor
• 19:38have these abrupt transitions
• 19:41between very highly cellular fascicular
• 19:44areas
• 19:45and slightly less cellular areas
• 19:47with a scant myxoid stroma,
• 19:49typically with this perivascular
• 19:52hypercellularity
• 19:53or condensation
• 19:54of the cells
• 19:55around the blood vessels.
• 19:57And if you see classic
• 19:58histology, that can be very
• 20:00helpful,
• 20:00but many cases don't look
• 20:02as good as this.
• 20:05Mentioning another chromatin remodeling complex,
• 20:08this is the polychrome repressive
• 20:10complex two.
• 20:12PRC two has multiple subunits,
• 20:14as you can see in
• 20:15this this nice diagram,
• 20:17including SUS12 and EED.
• 20:20And
• 20:20we've known for quite some
• 20:22time
• 20:23that this complex
• 20:25is a methylating enzyme that's
• 20:27recruited to chromatin
• 20:29to trimethylate
• 20:30histone h three
• 20:32at the lysine twenty seven
• 20:33residue.
• 20:35And we call that moiety
• 20:37h three k twenty seven
• 20:39m e three.
• 20:40And if you say it
• 20:41under your breath a few
• 20:42times, you'll learn to repeat
• 20:44it very easily.
• 20:45H three k twenty seven
• 20:47m e three,
• 20:48and now you know my
• 20:49iPhone password.
• 20:52So this modification
• 20:54is a way that we
• 20:55kinda keep our transcription in
• 20:57check,
• 20:57but modifications or dysregulation
• 21:00can lead to cancer of
• 21:02various types.
• 21:03And in two thousand fourteen,
• 21:05several groups discovered
• 21:07that mutations
• 21:09in one of the genes
• 21:10that encodes PRC2
• 21:13are found in most examples
• 21:16of high grade malignant peripheral
• 21:17nerve sheath tumor.
• 21:19And these alterations
• 21:21either in SUS12
• 21:22most often or sometimes an
• 21:24EED
• 21:25lead
• 21:26to inactivation
• 21:27of the enzyme.
• 21:29So PRC2
• 21:30can no longer trimethylt histone
• 21:32three
• 21:33at lysine twenty seven.
• 21:35And remarkably, there were already
• 21:37commercial antibodies available
• 21:39that only recognized histone three
• 21:42when it has
• 21:43lysine twenty seven trimethylation.
• 21:45So we're looking for loss
• 21:47of this normal trimethylation mark
• 21:50by immunoskechemistry.
• 21:51And this works best
• 21:53in intermediate to high grade
• 21:55examples, which are the ones
• 21:56we encounter most often
• 21:58in the sporadic setting in
• 21:59our clinical practice and surgical
• 22:01pathology.
• 22:02In low grade MPNST, we're
• 22:04really trying to decide is
• 22:05it malignant
• 22:06in a patient with NF1
• 22:07who's gets a atypical neurofibroma
• 22:10in transformation.
• 22:11We're not trying to decide
• 22:12is it a nerve sheath
• 22:13tumor.
• 22:15The specificity
• 22:16is also excellent.
• 22:18Almost all the tumors you
• 22:19would consider in the differential
• 22:20diagnosis
• 22:22show consistent
• 22:23normal
• 22:24staining with this antibody.
• 22:26The only two exceptions are
• 22:28dedifferentiated liposarcoma
• 22:32melanoma. That's a very small
• 22:34percentage.
• 22:35As you many of you
• 22:36know, Ddiff liposarq has MDM
• 22:38two amplification.
• 22:39So that's an easy way
• 22:41to make that diagnosis.
• 22:42And spindle cell melanomas arise
• 22:44in the skin. They spread
• 22:45to lymph nodes.
• 22:47They're strongly and diffusely SOX10
• 22:49and S100 positive in most
• 22:50cases. So it's usually not
• 22:52really a problem
• 22:53in differential diagnosis for a
• 22:55deep
• 22:56seated malignant spindle cell neoplasm
• 22:58with limited or no
• 23:00Sox ten expression.
• 23:01So this is what it
• 23:02looks like. It looks just
• 23:03like a tumor suppressor protein
• 23:05loss, but we're looking for
• 23:07we're seeing a loss of
• 23:08the trimethyl mark. And it
• 23:10works beautifully. We now use
• 23:12this daily in my practice,
• 23:14in our laboratory.
• 23:15And this was a case
• 23:16that would be impossible to
• 23:18recognize.
• 23:19It just looks like a
• 23:20fibrosarcoma
• 23:21like
• 23:22fascicular spindle cell malignant neoplasm,
• 23:25but we know this is
• 23:26MPNST because it arose in
• 23:27a patient with NF one.
• 23:30And one of the tumors
• 23:31that is most difficult to
• 23:33distinguish is synovial sarcoma.
• 23:35These tumors have normal staining
• 23:38for the trinephile antibody.
• 23:45So now we're gonna shift
• 23:46and talk a little bit
• 23:48about
• 23:49looking for protein correlates of
• 23:51gene fusions
• 23:53as predictive testing for targeted
• 23:55therapies.
• 23:56You all know about ALK.
• 23:58This is a very
• 24:00exciting topic in
• 24:02oncology, medical oncology,
• 24:04and ROS1, which is a
• 24:05very similar receptor tyrosine kinase
• 24:08gene.
• 24:09And about four percent of
• 24:11lung adenocarcinomas
• 24:12have gene fusions involving ALK.
• 24:14One percent have ROS1 fusions.
• 24:17And these alterations are much
• 24:18more common in inflammatory
• 24:20myofibroblastic
• 24:21tumors,
• 24:23within which they were identified
• 24:25before they were in lung
• 24:26cancer.
• 24:28This
• 24:30gene was named after
• 24:32lymphoma, as you all know.
• 24:34This is the anaplastic lymphoma
• 24:36kinase.
• 24:37Way back in nineteen ninety
• 24:39four,
• 24:40Morris and Look
• 24:41discovered
• 24:42that ALK fusions
• 24:44are a very common driver
• 24:46in a very uncommon
• 24:48aggressive
• 24:50t cell non Hodgkin lymphoma
• 24:52that's now called anaplastic large
• 24:54cell lymphoma,
• 24:55ALK rearranged.
• 24:57It was two thousand and
• 24:58seven that Souda and colleagues
• 25:00from Japan
• 25:01discovered the first gene fusions
• 25:03in lung cancer involving ALK.
• 25:05This was the EML4 ALK
• 25:07fusion.
• 25:08Many of these cases are
• 25:09mucinous
• 25:11or poorly differentiated
• 25:12adenocarcinomas
• 25:13with a signet ring cell
• 25:15component, although not invariably. They
• 25:17can have a range
• 25:18of histology within the spectrum
• 25:20of adenocarcinomas.
• 25:23And then remarkably, in only
• 25:24three years between that publication
• 25:27and the publication of this
• 25:28first phase one clinical trial,
• 25:31this group,
• 25:33predominantly from Massachusetts General Hospital
• 25:35in Boston,
• 25:36showed us that we can
• 25:38target,
• 25:39this
• 25:40fusion
• 25:41in advanced lung adenocarcinomas
• 25:44with ALK gene rearrangements
• 25:46as a very effective way
• 25:47of treating these unfortunate patients.
• 25:50And I just have to
• 25:51give a shout out to
• 25:52my friend, John Iefrady.
• 25:53The senior author is a
• 25:55molecular pathologist.
• 25:57I actually trained him in
• 25:58surgical pathology. He was a
• 25:59first year when I was
• 26:00a second year. John likes
• 26:01to say it didn't take
• 26:02because he doesn't do surgical
• 26:04pathology,
• 26:05but he's a brilliant molecular
• 26:06pathologist who actually developed Archer
• 26:09in his research lab,
• 26:11and he was a senior
• 26:12in this paper.
• 26:14And you can see from
• 26:15the waterfall plot, this had
• 26:17really dramatic benefit to these
• 26:18patients.
• 26:19I'm sure many of you
• 26:20know that crizotinib
• 26:21alone doesn't last very long.
• 26:24Patients develop secondary resistance mutations
• 26:27within the kinase domain of
• 26:29ALK, but now we have
• 26:30many different
• 26:31targeted therapeutics,
• 26:33second, third, fourth generation,
• 26:35and this has really transformed
• 26:37thoracic oncology practice, not only
• 26:39against ALK, but against ROS1
• 26:42and many other,
• 26:43kinase alterations.
• 26:47Beyond the tumors I've mentioned,
• 26:48we now know that ALK
• 26:49fusions are found in a
• 26:50wide range of tumor types.
• 26:53Some are highly aggressive,
• 26:54Some of them are usually
• 26:56indolent, and some of them
• 26:58are entirely benign.
• 26:59Even in these very trivial
• 27:01benign skin tumors that we
• 27:03call
• 27:04epithelioid
• 27:05fibrocystiocytomas.
• 27:06And they can have the
• 27:07exact same fusions
• 27:09as highly aggressive cancers.
• 27:13The ALK fusion partner sometimes
• 27:15results in a distinctive pattern
• 27:18by immunohistochemistry
• 27:19down the microscope, which is
• 27:20really quite amazing.
• 27:22The most common fusion in
• 27:24anaplastic large cell lymphoma with
• 27:26nucleophosmin
• 27:27results in this combined
• 27:29cytoplasmic
• 27:30and nuclear pattern.
• 27:32There's a very distinctive pattern
• 27:34we see in an aggressive
• 27:36variant of inflammatory
• 27:38myofibroblastic
• 27:39tumor
• 27:40that's called epithelioid
• 27:42inflammatory
• 27:43myofibroblastic
• 27:44sarcoma,
• 27:45and that is a nuclear
• 27:46membrane pattern of ALK.
• 27:48And this is all based
• 27:49on what the fusion partner
• 27:51is. Many of the other
• 27:52fusions result in a cytoplasmic
• 27:55pattern of staining.
• 27:57This is ALCL with nucleophosmin.
• 27:59You can see that kind
• 28:00of complex pattern of localization
• 28:03within the cells.
• 28:05This is lung adenocarcinoma
• 28:07with EML4 ALK fusion, a
• 28:09granular
• 28:10pattern of cytoplasmic staining.
• 28:12And this is that aggressive
• 28:13variant of IMT
• 28:15where the Ran binding protein
• 28:18two fusion
• 28:19brings ALK to the nuclear
• 28:21pore complex
• 28:22and it looks like this
• 28:23really distinctive, beautiful
• 28:25nuclear membrane pattern.
• 28:29For clinical practice, it's also
• 28:31very important to know that
• 28:33which antibody
• 28:34clone you're using is important
• 28:37in terms of what you're
• 28:38using the application for. The
• 28:40ALK one clone we've had
• 28:42for more than thirty years,
• 28:43and it works beautifully
• 28:45for lymphoma diagnosis,
• 28:47but it's rather insensitive. It's
• 28:49not a very,
• 28:50powerful antibody
• 28:52and
• 28:53lung adenocarcinomas
• 28:55with ALK fusions actually have
• 28:56a pretty low level of
• 28:57protein expression.
• 28:59So ALK one is usually
• 29:01negative, and it doesn't help
• 29:02you as a screen for
• 29:04EML four or other ALK
• 29:06gene fusions.
• 29:08In contrast,
• 29:09in the last ten years
• 29:10or so, there have been
• 29:11two
• 29:12highly sensitive anti ALK antibodies
• 29:15developed,
• 29:16which are beautiful surrogates for
• 29:17the fusions.
• 29:19And now they can be
• 29:20used as a standalone test
• 29:22to identify what patients with
• 29:24lung adenocarcinoma
• 29:26can be treated with targeted
• 29:28therapies.
• 29:30Inflammatory
• 29:31myofibroblastic
• 29:32tumors are kind of in
• 29:33the middle. The level of
• 29:34expression is usually
• 29:36sufficient
• 29:37for ALK1 to detect the
• 29:39ALK in IMTs,
• 29:41but you'll miss about five
• 29:42percent of cases that way.
• 29:44The newer antibodies, you'll pick
• 29:46up all the ALK fusions.
• 29:49And we were very lucky.
• 29:50We were able to piggyback
• 29:52along with that really
• 29:53fundamental,
• 29:55like
• 29:56practice changing paper on lung
• 29:58adenocarcinoma
• 29:59with crizotinib
• 30:01that I mentioned.
• 30:02Because of that paper, we
• 30:03could publish a case report,
• 30:05snuck into the same issue
• 30:07of the New England Journal
• 30:08of Medicine,
• 30:09the first patient with one
• 30:10of these aggressive variants of
• 30:12IMT
• 30:13who had a really dramatic
• 30:15response to crizotinib.
• 30:16And this was from this
• 30:18study.
• 30:19My former colleague, James Butrinsky,
• 30:22who's a medical oncologist specializing
• 30:25in sarcoma,
• 30:26treated this patient
• 30:27who had a pelvic
• 30:29epithelioid inflammatory myofibroblastic
• 30:32sarcoma.
• 30:33It recurred locally with multifocal
• 30:35disease, including
• 30:37a metastasis of the liver.
• 30:39The patient went on therapy.
• 30:40After three months, there was
• 30:41a dramatic partial response.
• 30:43Then a surgical oncologist
• 30:46by the name of Monica
• 30:47Bertinoli,
• 30:48who actually moved on to
• 30:49become the head of the
• 30:50NCI,
• 30:51and then she was the
• 30:52head of the NIH.
• 30:53So my claim to fame
• 30:54is I briefly worked with
• 30:55Monica, who's a brilliant
• 30:57scientist
• 30:58and a
• 31:00surgical oncologist.
• 31:01She did debulking of this
• 31:03patient,
• 31:03and he survived for more
• 31:05than ten years on oral
• 31:06crizotinib therapy after this. So
• 31:08this is not only transformative
• 31:10for patients with carcinomas,
• 31:12but there are some
• 31:15aggressive
• 31:16mesenchymal tumors
• 31:17that are targetable because of
• 31:19these alterations.
• 31:22And just a few words
• 31:23about ROS1. So this is
• 31:24ROS1 immunohistochemistry
• 31:26in one of the lung
• 31:27adenocarcinomas
• 31:28with a ROS1 fusion.
• 31:30But unfortunately, the antibody we
• 31:32have against ROS1
• 31:34is highly sensitive,
• 31:36but the specificity
• 31:37is less than the ones
• 31:40against ALK.
• 31:42So it's a very good
• 31:43screen,
• 31:44but we can't use it
• 31:45as a standalone predictive test
• 31:47for lung adenocarcinoma.
• 31:49So in our practice, we
• 31:50do run ALK and ROS1
• 31:52as routine
• 31:54screening for lung adenocarcinomas.
• 31:56If ALK is positive, that's
• 31:58enough information for treatment.
• 32:00But ROS1 requires
• 32:01confirmation
• 32:03by gene fusion testing or
• 32:05FISH.
• 32:08One other example I'm just
• 32:09gonna mention because it's been
• 32:10such a hot topic in
• 32:12oncology
• 32:13and that's NTRK, which I'm
• 32:14sure many of you know
• 32:15about.
• 32:16Three different NTRK genes result
• 32:18in three different proteins.
• 32:20One, two, and three becomes
• 32:22A, B, and C. And
• 32:23we have antibodies available
• 32:25that recognize a conserved sequence
• 32:28in all three of these
• 32:30Turk proteins.
• 32:31It's a very nice antibody.
• 32:33And why do we care
• 32:34about NTRK? Because very similar
• 32:36to what we saw in
• 32:38ALK prearranged lung cancers,
• 32:41NTRK,
• 32:42tumors with NTRK fusions that
• 32:44are very aggressive
• 32:46can be effectively treated
• 32:48with larotrectinib
• 32:49and other NTRK inhibitors. And
• 32:51in fact, in this first
• 32:53paper that was published almost
• 32:55eight years ago now,
• 32:57the waterfall plot looks even
• 32:58better than crizotinib,
• 33:00And as you'll notice from
• 33:01the top of this graph,
• 33:03this was agnostic to tumor
• 33:05type. So we had carcinomas,
• 33:08GIST,
• 33:10salivary gland tumors,
• 33:11melanoma,
• 33:13all a couple of sarcomas.
• 33:16Irrespective of the tumor type.
• 33:18If it was metastatic
• 33:20malignancy with an NTRK fusion,
• 33:22they had really good responses
• 33:24in most cases.
• 33:26The antibody works really well
• 33:28to identify tumors that are
• 33:30defined by NTRK fusions.
• 33:33For example, in
• 33:35infantile fibrosarcoma,
• 33:37this antibody, the Pentract antibodies
• 33:40become a very easy way
• 33:41to make the diagnosis.
• 33:43The same thing goes for
• 33:44secretory carcinomas of the breast
• 33:46or salivary gland or rarely
• 33:48in skin and other sites.
• 33:50These have,
• 33:51NTRK fusions as a defining
• 33:53feature and immunohistochemistry
• 33:55works beautifully. You see the
• 33:57nuclear staining
• 33:58in these secretory
• 33:59carcinomas.
• 34:01And several groups have actually
• 34:03evaluated,
• 34:05PennTrak immunostochemistry
• 34:07as a surrogate for NTRK
• 34:08fusions
• 34:09in tumor types where the
• 34:11NTRK rearrangements
• 34:12are much less common.
• 34:14And as I'll show you
• 34:15in a minute, sometimes these
• 34:16are very, very rare events.
• 34:18Many of these were led
• 34:19by Jacqueline Heckman from New
• 34:21York.
• 34:22Really beautiful papers,
• 34:24trying to decide
• 34:25whether we could use the
• 34:27antibody as a surrogate.
• 34:29And as you see from
• 34:30the left side of this
• 34:31panel,
• 34:32unfortunately,
• 34:33in many cancers that are
• 34:34much more common,
• 34:35the fusions are really rare,
• 34:38zero point one percent, zero
• 34:39point three percent of tumors.
• 34:42And Lynette and I did
• 34:43this
• 34:44prospective,
• 34:45very small screening study a
• 34:47couple of years ago
• 34:48just to kind of test
• 34:49out this hypothesis.
• 34:51We were using all the
• 34:52cases for which we were
• 34:53already doing
• 34:55testing for lung cancer,
• 34:57endometrial
• 34:58cancer, colon cancer for MMR,
• 35:00and we used one antibody
• 35:02to do a pen track
• 35:03over the course of, like,
• 35:04six months. And we found
• 35:06that four out of almost
• 35:08five hundred cases were positive.
• 35:10But by testing, only one
• 35:12of these had an NTRK
• 35:13fusion.
• 35:14So clearly, this is not
• 35:15a great method to screen,
• 35:17and it probably is not
• 35:18gonna solve
• 35:19the problem to identify these
• 35:21very rare NTRK rearranged aggressive
• 35:24cancers.
• 35:25And if you look at
• 35:26this other study by Rosen
• 35:28from Clinical Cancer Research
• 35:30from twenty twenty,
• 35:32you see that the sensitivity
• 35:33and specificity
• 35:34are imperfect.
• 35:35And that's really the problem.
• 35:37The sensitivity
• 35:38for NTRK3 is almost eighty
• 35:40percent, so you're gonna miss
• 35:42twenty percent of cases.
• 35:44And at the same time,
• 35:46the specificity overall is only
• 35:48about eighty percent.
• 35:49So, unfortunately, it looks like
• 35:51immuno stick chemistry doesn't save
• 35:52us as a screen for
• 35:54these patients and you really
• 35:55have to do
• 35:56fusion testing, which obviously in,
• 35:59resource poor settings, is it
• 36:01possible?
• 36:01So it's still a problem
• 36:03that we're trying to, figure
• 36:04out.
• 36:07Next, I'm going to show
• 36:08you a couple of really
• 36:09nice examples of mutant oncoprotein
• 36:12specific antibodies.
• 36:13We've had these for many
• 36:15years. Some of them are
• 36:16newer and some are older.
• 36:20I suspect many of you
• 36:21know about the IDH1
• 36:23R132H
• 36:25mutation specific antibody.
• 36:27This has become routine practice
• 36:29for diagnosis of gliomas.
• 36:31It's really a remarkably
• 36:33powerful way
• 36:35for classification of gliomas and
• 36:37prognostication.
• 36:38Fortunately,
• 36:39it's the most common IDH
• 36:41mutation by far in glial
• 36:43neoplasms.
• 36:44Only about six or seven
• 36:46percent
• 36:47of gliomas have mutations
• 36:49in IDH one or IDH
• 36:50two that are different amino
• 36:52acids substitutions.
• 36:54And I'm gonna show you
• 36:55a couple other nice examples
• 36:56here. So this is, this
• 36:58is the first example I
• 36:59just mentioned, the high grade
• 37:00glioma,
• 37:01and you can see that
• 37:02really beautiful immunoreactivity.
• 37:05If IDH1 is wild type,
• 37:06it's entirely negative.
• 37:08And this is done now
• 37:09every day in neuropathology
• 37:11practice
• 37:12everywhere,
• 37:13where you're diagnosing glial neoplasms.
• 37:17To show you one other
• 37:18nice example,
• 37:19I'm gonna just present this
• 37:20as an unknown.
• 37:22This is from an elderly
• 37:23man who presented with small
• 37:24bowel obstruction.
• 37:26When the surgeon went in
• 37:27and took out the segment
• 37:28of small bowel,
• 37:30you can see why there
• 37:31was this very large
• 37:33ulcerated mass that was kinking
• 37:35the bowel,
• 37:36growing into the lumen,
• 37:38involving the submucosa and the
• 37:39mucosa.
• 37:41And if you look at
• 37:41high power at this tumor,
• 37:44it has a really undifferentiated
• 37:46appearance,
• 37:47a sheet
• 37:48of epithelioid
• 37:49or oval cells with abundant
• 37:51pale cytoplasm
• 37:53and a very high mitotic
• 37:54rate.
• 37:55So now we're dealing with
• 37:57kind of the classic problem.
• 37:58Here's a very poorly differentiated
• 38:01malignancy.
• 38:02What could it be? Is
• 38:03it carcinoma, melanoma, sarcoma, something
• 38:06else?
• 38:07We do immunohistochemistry
• 38:09and everything is negative.
• 38:11So we're stuck.
• 38:12We have an undifferentiated
• 38:14malignant neoplasm,
• 38:16which is the oncologist's favorite
• 38:18diagnosis.
• 38:19When you make the diagnosis,
• 38:21they're totally fine. They know
• 38:23what to do. That's a
• 38:23joke because it's a really
• 38:25big problem.
• 38:27But in this case, we
• 38:28did one other marker,
• 38:30which was the BRAF V600E
• 38:32specific antibody.
• 38:34And as you can see,
• 38:35it's beautifully positive in the
• 38:36cytoplasm of these tumor cells.
• 38:39And this is in fact
• 38:40metastatic dedifferentiated
• 38:41melanoma.
• 38:43And what the surgeon didn't
• 38:44know is this patient actually
• 38:45had a history
• 38:47of a locally advanced melanoma,
• 38:48like fifteen years ago, I
• 38:50think, on the back, but
• 38:51nobody knew, but we're able
• 38:53to solve it by this
• 38:54antibody.
• 38:55And this is a relatively
• 38:56new
• 38:57discovery
• 38:58that there are a small
• 39:00subset
• 39:01of melanomas, usually metastatic but
• 39:04occasionally primary,
• 39:05that are entirely negative for
• 39:08all the lineage markers. They
• 39:09don't express any neural crest
• 39:11markers or any melanoma
• 39:13restricted antibodies.
• 39:15S100,
• 39:16Sox10, HMB, Melanate, tyrosines, they're
• 39:18all negative.
• 39:19And that's really difficult. This
• 39:21is a challenging diagnosis.
• 39:23If they're
• 39:24entirely undifferentiated
• 39:26from the beginning, then we
• 39:27call them undifferentiated
• 39:28melanoma.
• 39:29If they started off with
• 39:30a conventional melanoma
• 39:32and then it recurred or
• 39:33metastasized and lost its markers,
• 39:35we call
• 39:36them dedifferentiated
• 39:37melanoma.
• 39:39And one of the best
• 39:40early studies on this topic
• 39:41was from Abbas Agami from
• 39:43Germany
• 39:44who published a series of
• 39:46these tumors and really talked
• 39:47about
• 39:48how we can recognize them
• 39:49in surgical pathology, what testing
• 39:51we can do
• 39:52to help us.
• 39:54There was a nice genomic
• 39:56study by David Adams,
• 39:58from United Kingdom,
• 39:59which had very similar observations
• 40:01with some more genomic testing.
• 40:05And as I mentioned, until
• 40:06we had this concept,
• 40:09we're often left with this
• 40:10descriptive, horrible, unhelpful diagnosis.
• 40:13The clues to the diagnosis
• 40:15sometimes are the anatomic site.
• 40:18The axilla is a very
• 40:19rare site for sarcomas,
• 40:21but it's a very common
• 40:23site for the first presentation
• 40:25of metastatic melanoma to a
• 40:27lymph node.
• 40:29Sometimes these tumors have rhabdoid
• 40:31cytology. In fact,
• 40:33in adult malignant neoplasms,
• 40:36the most common tumor with
• 40:37rhabdoid cytoplasmic inclusions is metastatic
• 40:40melanoma.
• 40:41So that can be a
• 40:42clue.
• 40:43If you do genomic testing,
• 40:46you can find the characteristic
• 40:47mutations of melanoma.
• 40:49BRAF, NRAS, KIT, NF one,
• 40:52the other,
• 40:54alterations you see in melanoma.
• 40:56And if you use a
• 40:57large panel that in part
• 40:58of the algorithms, they will
• 41:00read out particular mutation signatures.
• 41:03You can find a UV
• 41:04signature and you know that
• 41:06it's metastatic melanoma and it's
• 41:08not a sarcoma.
• 41:10And I've shown you how
• 41:10we have mutation specific immunohistochemistry.
• 41:13Just a few more quick
• 41:15examples.
• 41:15This was an axillary mass
• 41:17in a patient with no
• 41:18known
• 41:19history,
• 41:20an undifferentiated
• 41:22sheet like
• 41:23epithelioid malignant neoplasm with open
• 41:26chromatin, small nucleoli, and lots
• 41:28of mitotic activity.
• 41:31The only stain that was
• 41:32positive was CAM five point
• 41:34two.
• 41:35This is a monoclonal antibody
• 41:36that recognizes keratin eight. It
• 41:38wasn't very strong, but there
• 41:40was a little bit of
• 41:40staining. So you could say,
• 41:41well, maybe it's an undifferentiated
• 41:43carcinoma.
• 41:44We got a little bit
• 41:45of keratin,
• 41:46but, unfortunately, that was wrong.
• 41:48It had PRAME expression
• 41:50as well as RAS q
• 41:51sixty one r, and this
• 41:53is a metastatic undifferentiated melanoma.
• 41:56And occasionally,
• 41:58metastatic melanomas can have some
• 41:59expression of keratins.
• 42:01It's a well known diagnostic
• 42:03pitfall for dometopathologists,
• 42:05but it's definitely a source
• 42:06of
• 42:07challenge for us in surgical
• 42:09pathology.
• 42:11One other example, this was
• 42:12an undifferentiated melanoma in an
• 42:14axillary lymph node. This tumor
• 42:17had some rhabdoid cytoplasmic
• 42:19conclusions.
• 42:20By meus to chemistry, it
• 42:21reacted with absolutely nothing
• 42:24except for BRAF V600E.
• 42:28One other nice example,
• 42:30Adrian Flanagan from London
• 42:32discovered in two thousand thirteen,
• 42:35the giant cell tumors of
• 42:36bone
• 42:37and chondroblastoma.
• 42:39These are the two
• 42:40very well known
• 42:42locally aggressive tumors that arise
• 42:44in the epiphysis of long
• 42:45bones,
• 42:46harbor mutations
• 42:48in histone H3 genes,
• 42:50the same gene that encodes
• 42:52the histone three protein
• 42:54that is trimethylated
• 42:56in
• 42:57loss of the trimethylation in
• 42:59malignant peripheral nerve root tumor.
• 43:01But in this case, there's
• 43:02amino acid substitutions.
• 43:04One particular substitution, g thirty
• 43:07four w,
• 43:08is found in eighty five
• 43:09to ninety percent of giant
• 43:11cell tumors of bone. There's
• 43:13a different substitution
• 43:14in chondroblastoma.
• 43:16In most cases, the diagnosis
• 43:18is pretty easy.
• 43:19Radiology is distinctive.
• 43:21We have mononuclear
• 43:22cells and huge numbers of
• 43:24osteoclasts that have very impressive
• 43:26numbers of nuclei.
• 43:28But if you're dealing with
• 43:29a small biopsy
• 43:31or the radiology is not
• 43:32entirely definitive,
• 43:34it's very useful to have
• 43:35a surrogate
• 43:36for these mutations.
• 43:38So now we use this
• 43:39histone H3G34W
• 43:42mutation specific antibody for giant
• 43:44cell tumor of bone.
• 43:46As you can see, the
• 43:47osteoclasts are not neoplastic, they're
• 43:49reactive.
• 43:50Even though we call it
• 43:51giant cell tumor, the giant
• 43:53cells are actually non neoplastic.
• 43:54It's the mononuclear cells that
• 43:56are positive,
• 43:57but it's really a very
• 43:58easy way to make the
• 43:59diagnosis.
• 44:01Very rarely,
• 44:03giant cell tumors of bone
• 44:04can transform
• 44:05to high grade sarcomas.
• 44:07They still retain
• 44:09the G34W
• 44:11or other histone H3
• 44:13substitutions.
• 44:17Now,
• 44:17mentioning again something that I
• 44:19that we talked about this
• 44:20morning in the resident slide
• 44:21seminar.
• 44:22This is a fusion specific
• 44:24antibody that was developed for
• 44:25synovial sarcoma.
• 44:29Synovial sarcoma is quite common.
• 44:31As you all know, it's
• 44:32one of the ten most
• 44:33common sarcomas.
• 44:35It's quite aggressive.
• 44:36The ten year survival is
• 44:38about fifty percent,
• 44:40and it's defined by a
• 44:41pathic mnemonic translocation
• 44:43between chromosomes X and eighteen.
• 44:46This results in fusions of
• 44:48genes that have changed their
• 44:50names a bit over the
• 44:51years.
• 44:51They're now named after the
• 44:53tumor type and the chromosome.
• 44:55So they're easy to remember.
• 44:57SS18
• 44:58and SSX. Synovial sarcoma
• 45:00on chromosomes eighteen and chromosomes
• 45:02X.
• 45:03We have three different variants,
• 45:05monophasic, which is a spindle
• 45:07cell neoplasm,
• 45:08biphasic that contains epithelial
• 45:10glands,
• 45:11and poorly differentiated that has
• 45:13round cell morphology
• 45:15that can be very difficult
• 45:17to distinguish
• 45:18from Ewing's sarcoma and other
• 45:19round cell sarcomas.
• 45:21And conventional markers are really
• 45:23not very helpful
• 45:25to diagnose synovial sarcoma.
• 45:27And until quite recently in
• 45:28most practices,
• 45:30we would confirm the diagnosis
• 45:32by sequencing or by FISH
• 45:34to look for the fusion
• 45:35or the SS18 rearrangement.
• 45:38And a couple of years
• 45:39ago, we worked with a
• 45:40colleague, Segal Kadoch, who's a
• 45:42scientist,
• 45:43at the Dana Farber Cancer
• 45:44Institute.
• 45:45When Segal was a post
• 45:46doc at Stanford before she
• 45:48joined the faculty at Dana
• 45:50Farber,
• 45:51she discovered the mechanism of
• 45:53transformation of cells
• 45:55by the synovial sarcoma fusion
• 45:56protein,
• 45:57and she's become one of
• 45:58the world experts in the
• 46:00SWISNF
• 46:01chromatin remodeling complex.
• 46:03And she
• 46:05recommended
• 46:06amino acid sequences to use
• 46:08to immunize animals
• 46:09to develop these monoclonal antibodies.
• 46:11And, essentially, we were just
• 46:13a test lab for these
• 46:14antibodies.
• 46:15The company she worked with
• 46:16kept sending us aliquots of
• 46:18the antibodies
• 46:19that worked beautifully in Western
• 46:20blot, but most of them
• 46:22didn't work at all. We
• 46:23couldn't optimize them to get
• 46:24any signal
• 46:25in formal and fixed paraffin
• 46:27embedded tissue administered chemistry.
• 46:29And eventually,
• 46:30in late
• 46:31nineteen,
• 46:33two thousand nineteen,
• 46:34we finally,
• 46:36uncovered two of their antibodies
• 46:37that worked beautifully.
• 46:39One of them is a
• 46:39fusion antibody that recognizes the
• 46:41sequence that crosses
• 46:43the break point between the
• 46:45two genes,
• 46:46which was ninety five percent
• 46:48sensitive and a hundred percent
• 46:49specific
• 46:50in this study that included
• 46:52a range of
• 46:53potential histologic mimics.
• 46:55And we also found one
• 46:56of their antibodies
• 46:58directed against a conserved sequence
• 47:00at the C terminus of
• 47:02all the SSX proteins was
• 47:04the opposite. Hundred percent sensitive
• 47:07and a little bit less
• 47:08specific.
• 47:09And it works beautifully.
• 47:11So this is genetic testing.
• 47:13We're looking for the fusion
• 47:14protein.
• 47:15And every single nucleus is
• 47:17positive because it's a genomic
• 47:19alteration
• 47:20in these cells.
• 47:21This is classic monophasic synovial
• 47:23sarcoma,
• 47:24very cellular spindle cell neoplasm
• 47:26that, as one of the
• 47:27residents described this morning,
• 47:29very little cytoplasm. They're almost
• 47:31overlapping nuclei.
• 47:33They often have these little
• 47:34wiry collagen
• 47:35bundles between the cells.
• 47:37Beautiful reactivity
• 47:39with the
• 47:40SS18SSX
• 47:42fusion specific antibody.
• 47:44And this is poorly differentiated
• 47:45synovial sarcoma
• 47:47that can look very similar
• 47:48to other high grade round
• 47:50cell sarcomas.
• 47:52And the fusion antibody is,
• 47:53again, beautifully positive.
• 47:57And finally, before I end,
• 47:58I'm just gonna talk about
• 48:00using immunohistochemistry
• 48:01to screen
• 48:02for familial predisposition syndromes.
• 48:05You all know about mismatch
• 48:07repair protein immunohistochemistry.
• 48:09We've been doing this for
• 48:10decades.
• 48:11This is a very useful
• 48:12way to screen for Lynch
• 48:13syndrome
• 48:14in colorectal endometrial adenocarcinoma,
• 48:18upper urinary tract adenocarcinomas,
• 48:20for example,
• 48:21the tumor types that are
• 48:22most common in patients with
• 48:23Lynch syndrome with germline mutations.
• 48:26And now that's become a
• 48:27huge area of surgical pathology
• 48:29practice.
• 48:30Now that we have
• 48:31immune checkpoint inhibitor therapy, which
• 48:33is most effective
• 48:35for cancers that have mismatch
• 48:37repair deficiency
• 48:39outside of Lynch syndrome as
• 48:40well when they're sporadically inactivated
• 48:42either by methylation of the
• 48:44MLH1 promoter or by somatic
• 48:46mutations, for example.
• 48:48But we have a couple
• 48:49other examples of how we
• 48:50can use antibodies
• 48:52to screen for the potential
• 48:55alterations that that belies some
• 48:57of these other
• 48:59tumor syndromes.
• 49:00The one I'm gonna mention
• 49:02is succinate dehydrogenase.
• 49:04We've known for quite some
• 49:05time that patients with Kearney
• 49:08Triad
• 49:09described by Aden Kearney at
• 49:10the Mayo Clinic in nineteen
• 49:11seventy seven,
• 49:13developed these three distinctive tumor
• 49:15types,
• 49:16unusual form of gastric
• 49:18gastrointestinal
• 49:19stromal tumor or GIST,
• 49:21extra adrenal paraganglioma,
• 49:23and pulmonary chondroma.
• 49:25And we didn't really know
• 49:27why for many years. This
• 49:28is not a familial syndrome.
• 49:30It's sporadic, and it usually
• 49:31affects young women.
• 49:34About ten years ago, it
• 49:35was discovered
• 49:36that this is caused by
• 49:38hypermethylation
• 49:39of the SDHC promoter.
• 49:42There is a much more
• 49:43common
• 49:44familial paraganglioma
• 49:46syndrome, which is responsible for
• 49:48about twenty percent of all
• 49:49paragangliomas,
• 49:51which is caused by germline
• 49:53mutations
• 49:54in either SDHB,
• 49:56SDHC, or SDHD.
• 49:58This has very high penetrance,
• 50:01and you could use immunohistochemistry
• 50:02as a screen as I'll
• 50:03come back to in a
• 50:04few minutes.
• 50:05One other syndrome is called
• 50:07Kearney Stratakis, which is very
• 50:09similar to Kearney triad, except
• 50:11these patients do not develop
• 50:13pulmonary chondromas.
• 50:15And this is essentially just
• 50:16a variant of the familial
• 50:17paraganglioma
• 50:18syndrome where they also develop
• 50:20these distinctive gists.
• 50:23I'm sure many of you
• 50:24know in surgical pathology
• 50:26that we really can't predict
• 50:28which paragangliomas
• 50:29are going to be malignant.
• 50:31That's been something we've tried
• 50:32to deal with for many
• 50:33decades without a lot of
• 50:34success.
• 50:36It turns out
• 50:37the most
• 50:38powerful predictor of malignancy in
• 50:41paragangliomas
• 50:42is SDHB
• 50:43mutation,
• 50:45germline mutations.
• 50:46Patients with this syndrome who
• 50:48have that particular alteration
• 50:50have a risk of metastasis
• 50:51that's about fifty percent.
• 50:53There's also a very unusual
• 50:55kind of narrow anatomic distribution
• 50:58based on what the underlying,
• 51:01genetic alteration is, as you
• 51:03can see from this table.
• 51:07So it turns out that
• 51:09irrespective of which gene is
• 51:11mutated
• 51:12or if SDHC has promoter
• 51:14hypermethylation,
• 51:16the SDHB
• 51:17part of this enzyme complex
• 51:19gets degraded.
• 51:20It requires stability of all
• 51:21the elements in order for
• 51:23it to sit there.
• 51:24So you can use one
• 51:25antibody for SDHB
• 51:27as a surrogate for an
• 51:29inactivation of the complex.
• 51:30So this has now become
• 51:31a very easy way
• 51:33to screen
• 51:34for the familial paraganglioma
• 51:36syndrome.
• 51:37The first nice paper demonstrating
• 51:39how you might do this
• 51:40in clinical practice
• 51:41was published in Lancet Oncology
• 51:43by a group from the
• 51:44Netherlands,
• 51:45and this has become
• 51:47standard of practice for us
• 51:48now,
• 51:49in collaboration with, Justine Barletta,
• 51:51our head of endocrine pathology,
• 51:53and our colleagues at the
• 51:55genetics
• 51:56clinic at the Dana Farber
• 51:57Cancer Institute.
• 51:58Every patient with periganglioma
• 52:00or pheochromocytoma
• 52:02gets immunohistochemistry
• 52:03for SDHB as a screen
• 52:05to direct them for,
• 52:07germline testing.
• 52:09This is a pheochromocytoma
• 52:11of the adrenal gland with
• 52:12normal staining.
• 52:13It's this mitochondrial
• 52:15granular cytoplasmic pattern, as you
• 52:17can see here.
• 52:18This is an SDHB
• 52:20mutants paraganglioma.
• 52:22In contrast, the tumor cells
• 52:25show a complete loss of
• 52:26the normal cytoplasmic
• 52:27staining,
• 52:28whereas the endothelial
• 52:30cells
• 52:32and the sustentacular
• 52:33cells,
• 52:34those specialized Schwann cells that
• 52:36invest the individual
• 52:38zelbollen of the paraganglioma,
• 52:40show normal granular staining.
• 52:43As I mentioned, the gist
• 52:45that arise in the context
• 52:46of STH deficiency
• 52:48are really special.
• 52:50They have this multi nodular
• 52:52plexiform architecture,
• 52:54as you can see in
• 52:55the scanning images on the
• 52:56left and
• 52:57the intermediate power
• 52:58on the right, they're usually
• 53:00epithelioid
• 53:01and they have a very
• 53:03high rate of spreading to
• 53:04regional lymph nodes.
• 53:05About thirty percent of patients
• 53:08will have perigastric
• 53:09lymphometastases
• 53:11at the time of their
• 53:12partial gastrectomy.
• 53:13And this is in contrast
• 53:15to KIT mutant GIS,
• 53:16which spread to lymph nodes
• 53:18in about zero point one
• 53:19percent of cases. That's an
• 53:20incredibly rare event.
• 53:22These tumors
• 53:24very frequently metastasize
• 53:26to the perineal cavity or
• 53:28the liver, but they have
• 53:29a relatively indolent growth, and
• 53:31some patients can survive
• 53:33for years without therapy, without
• 53:36getting
• 53:37really sick. Unfortunately, many of
• 53:39them eventually succumb, but they
• 53:40can have a protracted course,
• 53:42which is totally different from
• 53:44KIT mutant GIS
• 53:45without targeted therapy. Once they
• 53:47metastasize,
• 53:48patients die of disease within
• 53:50about twelve to eighteen months.
• 53:52So totally different biology. They
• 53:53don't respond to
• 53:55kit inhibitors, obviously, because they're
• 53:57driven by a totally different
• 53:58pathway.
• 54:00And at the same time,
• 54:01we can't predict
• 54:03which patients are going to
• 54:04metastasize
• 54:05for our risk stratification criteria.
• 54:08So this is a special
• 54:09tumor type which is being
• 54:11separated from the rest of
• 54:12the GISTS
• 54:13in the twenty twenty six
• 54:15WHO classification,
• 54:17both the digestive disease volume
• 54:18and the soft tissue volume
• 54:20which will come out later
• 54:21this year.
• 54:23Anthony Gill from Sydney, Australia
• 54:25first showed us how we
• 54:26can use
• 54:27immunohistochemistry
• 54:28for SDHB to identify these
• 54:30tumors.
• 54:31This is very easy to
• 54:32apply. It works very well.
• 54:34And similar to what we
• 54:35saw with the paragangliomas,
• 54:37we have a complete loss
• 54:39of normal cytoplasmic
• 54:40staining
• 54:41in the tumor cells. And
• 54:43again, we have an internal
• 54:44control,
• 54:45endothelium
• 54:46or
• 54:47mucosa
• 54:48or the smooth muscle of
• 54:50the muscularis propria.
• 54:53So my closing comments,
• 54:55I've shown you we have
• 54:56this rapid evolution of a
• 54:58new generation of markers,
• 55:00femininistic chemistry,
• 55:01that are making it really
• 55:03easy to identify
• 55:04the molecular genetic alterations that
• 55:07define various cancer types,
• 55:09helping us in our surgical
• 55:10pathology practice,
• 55:12replacing molecular genetic testing in
• 55:14many cases.
• 55:16In addition,
• 55:17this is a nice form
• 55:18of rapid and inexpensive
• 55:20predictive testing for targeted therapies.
• 55:23And finally, just showed you
• 55:24one example
• 55:25of how we can use
• 55:26this technique
• 55:27as broad screening for familial
• 55:29cancer predisposition
• 55:31syndromes.
• 55:32Thank you again for inviting
• 55:34me to share this session
• 55:36with you today, and I'm
• 55:37happy to answer any questions.
• 55:52Yes. Great talk. Thank you.
• 55:55I just have a question
• 55:57about the the first example.
• 56:02Because when I was a
• 56:03resident,
• 56:04I was looking at this
• 56:05at the NIH, and I
• 56:06had no idea. I'm tired.
• 56:08This is tumor. I actually
• 56:10asked the attending
• 56:11who's trying to teach me
• 56:12already.
• 56:14That's all situated.
• 56:16Exactly. And,
• 56:18I I it's the specific
• 56:20form of pericyclinoma
• 56:22that's that's responsive to that,
• 56:25that marker, is it just
• 56:26present in the nose, or
• 56:27does that occur also for
• 56:29for Yeah. No. It's amazing.
• 56:31This is only sinonasal.
• 56:32So this is a very
• 56:33special tumor
• 56:35that only arises in the
• 56:36sinonasal tract. All the other
• 56:38parasitic tumors in the rest
• 56:39of the body have totally
• 56:41different
• 56:41genetics. Like, glomus tumors and
• 56:43some myoperacytomas
• 56:45have notch gene fusions.
• 56:47Totally different.
• 56:49It's really it's amazing. This
• 56:50is only sino nasal.
• 56:52And, you know, it's such
• 56:53a beautiful,
• 56:55you know,
• 56:56example of fleshing out all
• 56:58of these very, you know,
• 56:59rare things, but with very
• 57:01specific genetic alteration.
• 57:03Is it
• 57:04your sense that this is
• 57:06gonna continue, and and this
• 57:08is gonna play out building
• 57:09an?
• 57:10Yeah. And we we so
• 57:12the question is, you know,
• 57:13using chemistry
• 57:15to look for the genetic
• 57:16alterations, is this gonna kinda
• 57:17continue? And, you know, we
• 57:18still are working on more
• 57:20antibodies to identify fusions.
• 57:23I think we have a
• 57:24long way to go to
• 57:24make it useful, but, you
• 57:26know, many people kinda struggle
• 57:27with, well, in a practice
• 57:29that has very easy assets
• 57:30to to fusion testing by
• 57:32RNA NGS,
• 57:34should we just do that
• 57:34instead of developing these new
• 57:36antibodies? And it sort of
• 57:37depends on volume and your
• 57:39practice models.
• 57:41Some of the relatively common
• 57:42ones that we do see
• 57:43with some frequency, I think
• 57:45immunohistochemistry
• 57:46makes a lot of sense.
• 57:47Probably if you work at
• 57:48a cancer center that doesn't
• 57:49have as many of a
• 57:50particular group of tumors,
• 57:52sequencing is probably preferable.
• 57:54Yes. Yeah.
• 57:56Very nice talk. Thank you.
• 57:58This is slightly peripheral and
• 57:59maybe somewhat of a trivial
• 58:01question, but since you were
• 58:02the head of the, you
• 58:03know, that little while,
• 58:06how do you manage the
• 58:08logistics of this other than
• 58:09getting, you know, like a
• 58:10not your old woman to
• 58:11go in every day and
• 58:13hand singing stuff? Like, you're
• 58:14having all these additional
• 58:16things added in all Yeah.
• 58:19Yeah. So the question is
• 58:20sort of how do you
• 58:21logistically, how do you deal
• 58:22with this? And he was
• 58:24referring to our now retired
• 58:25head of hematopathology
• 58:27who liked to do her
• 58:28own pipetting and her own
• 58:29lab for hematopathology,
• 58:30Jerry Pincus, who's amazing, brilliant
• 58:33hematopathologist
• 58:34and one of the pioneers
• 58:35of immunohistochemistry,
• 58:36when they were all frozen
• 58:37sections, she was doing that
• 58:38as well. She's amazing.
• 58:41You definitely have to have
• 58:42a very good technical group.
• 58:44I'm very lucky. My lab
• 58:45supervisor,
• 58:46she's a an immunohistochemical
• 58:49technologist who'd been doing it
• 58:50for thirty years.
• 58:51She's really adept at optimizing
• 58:53antibodies. So
• 58:55I work with her to
• 58:56identify a commercially available source
• 58:59where it's a company that
• 59:00I pretty much trust that
• 59:01I think they've validated enough
• 59:02that I believe them,
• 59:04that it works in FFPE.
• 59:05Sometimes they're still lying and
• 59:07it's wrong, but often they
• 59:08work. So I'll order an
• 59:09antibody and I'll just have
• 59:11a hypothesis. I'll test it
• 59:13in a tumor with a
• 59:13known alteration.
• 59:15If it works, then I
• 59:16just do a small pilot
• 59:17with a resident or fellow.
• 59:18So they'll have a nice
• 59:19project to present at one
• 59:20of our use cap meetings
• 59:22and a nice proof of
• 59:23principle paper. And then I
• 59:25bring it on as a
• 59:26clinical test. But it definitely
• 59:27takes a lot of legwork,
• 59:29takes technical expertise, it takes
• 59:32energy for a trainee or
• 59:34a junior colleague to
• 59:36pull a bunch of cases
• 59:37with known genetics and the
• 59:39differential diagnosis,
• 59:40you know, so that you
• 59:41can make sure
• 59:42that the specificity is enough
• 59:44for it to be a
• 59:45valuable in practice.
• 59:47It is not
• 59:48it is not straightforward unless
• 59:49you have the infrastructure to
• 59:51support that. And then I
• 59:52guess I'm curious on the
• 59:54latter half. Once you've established
• 59:56Yeah. How do you get
• 59:56it into your workflows? You're
• 59:58getting the turnaround
• 59:59of such a large
• 01:00:00panel panel of tests so
• 01:00:02many places have been. Yeah.
• 01:00:04The question is about turnaround
• 01:00:05and bringing it into the
• 01:00:06clinical armamentarium.
• 01:00:08We we only do one
• 01:00:09run a day still.
• 01:00:12Yeah. We we do one
• 01:00:13immunosy chemistry run a day.
• 01:00:15Every morning, they come out
• 01:00:16at, like, two o'clock, and
• 01:00:17we just have enough machines
• 01:00:19that we can run a
• 01:00:21bunch of tests.
• 01:00:22I don't know. It's like,
• 01:00:23you need a lot of
• 01:00:24machines. If you have a
• 01:00:25very high volume practice, we
• 01:00:26run probably seven hundred
• 01:00:29tests a day. It's a
• 01:00:30lot
• 01:00:31of lot of machines.
• 01:00:35Yes.
• 01:00:36The few statements are not
• 01:00:37the UPR and Yes. Not
• 01:00:39but the central plastic. It's
• 01:00:40very hard dealing with the
• 01:00:41setting.
• 01:00:44Yes. How do you handle
• 01:00:46this with this IFC?
• 01:00:48Yeah. I think you have
• 01:00:49to make sure that your
• 01:00:50sort of signal to noise
• 01:00:52ratio is is enough. And
• 01:00:54when it's a universally expressed
• 01:00:56protein, for example,
• 01:00:57if you really crank it
• 01:00:58as it were and you
• 01:00:59get a very strong signal
• 01:01:00and you're looking for loss,
• 01:01:02then you can feel confident
• 01:01:03there is true loss. I
• 01:01:04think the problem
• 01:01:05is when you've optimized an
• 01:01:07antibody
• 01:01:08and the signal's weak and
• 01:01:09everything, then it's very hard
• 01:01:11to decide if there's loss,
• 01:01:12especially as you said if
• 01:01:14it's cytoplasmic. So you have
• 01:01:15to make sure the signal's
• 01:01:17really strong. So we we
• 01:01:19actually use some antibody detection
• 01:01:21kits where the chemistry really
• 01:01:23boosts the signal even if
• 01:01:25it's not kinda standard practice
• 01:01:27when you're dealing with some
• 01:01:28antibodies that have pretty weak
• 01:01:30results.
• 01:01:31And some of them are
• 01:01:32the ones that we really
• 01:01:33like to use, like the
• 01:01:34highly sensitive ALK and ROS1
• 01:01:36and BRAF e six hundred
• 01:01:37d. The antibodies aren't that
• 01:01:38strong, so you really have
• 01:01:39to boost the signal so
• 01:01:41you can believe it that
• 01:01:42it when it's positive, it's
• 01:01:43real. And with loss, it's
• 01:01:45the same problem.
• 01:01:46So a lot of it's
• 01:01:47just kind of validation and
• 01:01:48optimization and feeling comfortable with
• 01:01:50the results.
• 01:01:51Yeah.
• 01:01:53Thanks.
• 01:01:54Thanks.
• 01:01:55I think I've seen those,
• 01:01:58platforms, like, moving form, code
• 01:02:00expedite to a ton of
• 01:02:01things that once it's Yeah.
• 01:02:02Even that color is. So,
• 01:02:04you know, you really have
• 01:02:05that, you know, talk sort
• 01:02:07of look. Yeah. So, like,
• 01:02:08doing all of these on
• 01:02:09one session.
• 01:02:11Do you think
• 01:02:12we're pretty close, somewhat close
• 01:02:14to doing that instead of
• 01:02:15our HC?
• 01:02:16Yeah. I mean, I so
• 01:02:17the question is about
• 01:02:19the technologies that are sort
• 01:02:20of multiplexing in really impressive
• 01:02:22ways, often using kind of
• 01:02:23digital,
• 01:02:25and artificial intelligence can be
• 01:02:27very supportive of that. You
• 01:02:28have people in your department
• 01:02:29who know a lot more
• 01:02:30about that than I do.
• 01:02:31I don't do any of
• 01:02:31that. I do like very
• 01:02:33old school immunosystem chemistry, and
• 01:02:35I don't really, I don't
• 01:02:36keep abreast of that technology
• 01:02:38enough to know how close
• 01:02:39we are to routine implementation.
• 01:02:41But you had a question
• 01:02:42or comment? Yes. I guess.
• 01:02:44Thanks. So everything you should
• 01:02:46list today was really fascinating,
• 01:02:47but it's all minor. Yes.
• 01:02:49The following that continuous expression
• 01:02:52Yes. Yeah. The question is,
• 01:02:54these are all binary, and
• 01:02:55that's exactly right. So for
• 01:02:57me, colorimetric,
• 01:02:59right field, immuno chemistry,
• 01:03:02unless it's binary,
• 01:03:03I don't really know how
• 01:03:04to do it. And I
• 01:03:05know that you are doing
• 01:03:07a lot of great techniques
• 01:03:08to try to get readouts
• 01:03:09that are much more granular
• 01:03:11along a spectrum. But for
• 01:03:13me, I think using these
• 01:03:15old school techniques,
• 01:03:17you can't really get very
• 01:03:19good gradations of intensity of
• 01:03:21staining. So I am really
• 01:03:22just looking for genetic alterations
• 01:03:24that are plus minus. And
• 01:03:26you definitely need other methods
• 01:03:28to be able to get
• 01:03:30a a much more kinda
• 01:03:32active range of expression of
• 01:03:33proteins.
• 01:03:35Yeah. And, obviously, your department
• 01:03:37has done a lot on
• 01:03:38this, and I I've read
• 01:03:39a lot of it, and
• 01:03:40it's very impressive, but I
• 01:03:41don't do that stuff.
• 01:03:44Yeah. Please.
• 01:04:07Yeah.
• 01:04:09Yeah. Yeah. The question is
• 01:04:10in in sort of resource
• 01:04:12poor settings
• 01:04:13that really are don't or
• 01:04:14might never have access to
• 01:04:16genomic testing of any kind,
• 01:04:18does do these forms of
• 01:04:19immunohistochemistry,
• 01:04:20is that a possibility? And
• 01:04:22again, I think that's
• 01:04:23this is still not inexpensive.
• 01:04:26So I know that in
• 01:04:27many countries, even immunohistochemistry
• 01:04:29is impossible. But, you know,
• 01:04:30for example, you know, we
• 01:04:31do a lot of outreach
• 01:04:32with some countries through the
• 01:04:34Partners in Health kind of,
• 01:04:35you know, organization, including, like,
• 01:04:37in in Rwanda.
• 01:04:39And we've helped set up
• 01:04:40a laboratory there where they
• 01:04:41do immunohistochemistry.
• 01:04:43We kinda help advise the
• 01:04:44panel and some of the
• 01:04:46relatively common sarcomas
• 01:04:48that they see in their
• 01:04:49population, they actually get some
• 01:04:50of these antibodies. So I
• 01:04:52think that certainly,
• 01:04:54you need to have some
• 01:04:55level of expertise by the
• 01:04:57pathologist, which is another big
• 01:04:58problem because
• 01:04:59not all countries have the
• 01:05:01kinds of training that the
• 01:05:02Western world does for
• 01:05:04diagnostic immunoistochemistry
• 01:05:05and surgical pathology. But, I
• 01:05:08mean, really, I hope that
• 01:05:09this is helpful for
• 01:05:11settings where they don't have
• 01:05:12any genetic or or, you
• 01:05:13know, molecular genetic testing.
• 01:05:18Well, thank you again so
• 01:05:19much for for the invitation.