Discovering A Monoblast Marker

April 08, 2022

Information

April 7, 2022

Yale Pathology Grand Rounds

Mina Xu, MD

ID7661

To CiteDCA Citation Guide

00:00Today's grand round speaker
00:02is Mina Shu or very own Mina.
00:06She needs no introduction.
00:09Actually, Mina has been here longer than I,
00:12but I'll introduce Mina for the
00:15benefit of the new folks Mina grew up
00:20in California and came to northeast
00:23for college at Harvard University,
00:27and then she went back for MD at.
00:30UCSF and she must have missed
00:34the seasons and the snow,
00:36so she came back to Yale
00:39to train in pathology.
00:41During her training,
00:42Mina rose to become chief resident and Mina
00:47showed an early interest in hematology.
00:50She spent a year doing research
00:54at Brigham and Women's Hospital.
00:59During her medical school and then after
01:03finishing her pathology residency at Yale,
01:07she went back to Brigham and Women's
01:11Hospital to do a clinical fellowship.
01:14And as minard's,
01:16if you logged in earlier you
01:19heard me say there was no room
01:22to do anything else but clinical
01:25work with voice extremely heavy.
01:30Fortuitously, you know was recruited
01:33right back at faculty position at
01:36Yale pathology, and we are very
01:39lucky to have Meena here with us.
01:43Mina has established a track record of
01:47pursuit of curiosity and excellence in
01:50all her endeavors and I would say all her
01:54endeavors because she won many grants,
01:57awards and scholarships.
01:59From early age, in fact,
02:02several in high school making it
02:05to Harvard University Dean's list.
02:09At Yale, an outstanding achievement
02:12in autopsy or world during residency.
02:15Several Chairmans challenge grant
02:18awards and a Gilleard foundation
02:21grant that is just to name a few.
02:25Mina is a consummate clinician and
02:28an excellent research collaborator.
02:31She has close to add publications
02:34to her credit. In addition,
02:37Mina is an outstanding teacher and mentor,
02:40and that's demonstrated by her various
02:44speaking engagements all over US,
02:47Canada, and also China.
02:49She has given short courses,
02:52interactive microscopy sessions and
02:55participated in workshops and symposia
02:59at annual meetings of the US and
03:03Canadian Academy of the Theology.
03:06And at other professional organizations.
03:09And her mentees,
03:11mostly residents and fellows,
03:13they have successfully presented
03:16abstracts at national and international
03:19meetings and published the manuscripts
03:22in impactful journals.
03:24So with no further ado,
03:27let me now present her work today.
03:31You know?
03:36Thank you so much mandjou for
03:39that really amazing introduction
03:42that I probably don't deserve,
03:45but I just wanted to say, you know,
03:47thank you also for asking me to do
03:50this talk when you first asked me,
03:52I wasn't sure which of the many
03:55cool stories that were involved
03:57in in heme path I should present,
03:59but I quickly decided to do this one
04:02because I think you would agree with me.
04:05And and probably malani as
04:08the current IHC director.
04:10That now is a good time to tell
04:13our trainees and students about the
04:15importance of discovering new markers
04:18and that we are not at the end of the
04:21IHC era but at the beginning and we
04:25will get even better during the next ten,
04:2920-30 years at investigating protein
04:32expression in human tissues.
04:34And then cancer in particular.
04:37So these are my COI disclosures.
04:40I don't think that any of this
04:42will impact our talk today.
04:44And this is my outline,
04:45so I wanted to do a case presentation of a
04:50patient recently seen this phantom menace,
04:53which is our counting of blast and
04:56blast equivalents on heme path.
04:58A New Hope or marker for AMOL,
05:01and then a few other projects
05:04I spun off the original one.
05:07And by the way,
05:08I am not actually a Star Wars geek.
05:11Some of my friends are and they.
05:15I have thought in their childhood
05:17that becoming a Jedi is a legitimate
05:20career choice,
05:21which I think is kind of funny
05:23because as a pathologist,
05:24sometimes I think we are living
05:27that dream because we're actually
05:30able to on a daily basis,
05:32force miniscule pieces of tissue
05:35to tell us their truth and that to
05:39me is amazing and quite Jedi like.
05:41So here's my books case presentation.
05:45This is a 58 year old woman with
05:47a 2 year history of chronic
05:50myelomonocytic leukemia,
05:51and I actually saw that initial CML
05:55diagnosis now with progressive cytopenias
05:57bone pain and circulating glass and
06:00she is admitted for the symptoms.
06:04So her peripheral blood showed 8% blasts.
06:07These are mono blastic cells.
06:12She also had degranulate poesis.
06:15So dysplasia of the granulocyte
06:17lineage with abnormal folding
06:19of the granulocytes and in other
06:21areas hypo granularity.
06:23She also had a monocytosis as is
06:25normally seen for her blood smear.
06:28These delicately folding cells
06:30with Gray blue cytoplasm.
06:32These are mature monos.
06:34They proceeded to do a
06:36bone marrow aspiration,
06:38but it was a dry tap,
06:39most likely because of the fibrosis
06:42in the core biopsy.
06:44This was the core biopsy in which you
06:46see it is hypercellular for her age and
06:49there is dysplasia in the megakaryocytes.
06:52You can also see a myeloid predominance
06:55with hardly any erythroid islands here.
06:59But at least there is maturation in the
07:02form of metamyelocytes granulocytes.
07:04So actually if you compare to
07:07her initial bone marrow biopsy,
07:09this area looks very similar
07:11to the initial one,
07:13but a little more hypercellular.
07:15However,
07:16about 30% of the bone marrow
07:19actually showed these
07:20foci of immaturity.
07:22So here I say immaturity because the
07:24cells have more dispersed chromatin,
07:27some distinct nucleoli.
07:29And you're having a lack
07:31of the mature granulocytes,
07:34so this is very worrisome,
07:35especially in this clinical context
07:39of transformation into AML.
07:42So can we treat this patient as AML now?
07:45Well, the definition of AML is
07:4820% loss in blood or bone marrow.
07:50So how can we reach that with
07:52our without our gold standard
07:53of the aspirin count?
07:57If the AML expressed our most
07:59utilized immunostain which is CD 34,
08:02we could demonstrate that on core biopsy,
08:05but without that marker.
08:06It would be very difficult to substantiate
08:09the cytologic blast count on core,
08:12and this was actually negative,
08:14as we knew from before.
08:17So just taking a step back and L is a
08:20genetically heterogeneous myeloid neoplasm,
08:23and while the FAAB classification is
08:25not employed in clinical use now,
08:28it is still the most adherent to
08:31myeloid differentiation status,
08:32so you'll see that it still comes into
08:34play in current research studies.
08:37The overall poor 5 year old survival
08:39for AML as well as the relapse
08:41rate continue to be a huge problem.
08:46You can see that some improvements
08:48have been made in the last couple
08:50of decades as seen here in Black is
08:542000 to 2006 and yellow is until
08:572011 and blue is the most current.
09:00This is from a Danish study that I'm
09:02using because it's one of the most recent,
09:04but those published from the
09:06US show similar statistics.
09:08You can see that there is still a long
09:10way to go for patients over the age of 60.
09:16In broad strokes, some of the
09:18major AML subtypes are classified
09:20according to cytogenetic findings
09:22that include balanced translocations.
09:25Molecular findings also
09:27help in risk stratification.
09:29Older age remains one of the most
09:31major risk factors for poor outcome.
09:36The foundation of treatment for
09:38new AML is induction chemotherapy
09:41followed by either consolidation,
09:43chemo or consideration toward
09:45allogeneic stem cell transplant.
09:48Newer therapeutic options include
09:50small molecule inhibitors, and in 2018,
09:53BCL 2 inhibitor venetoclax was
09:55approved as a combo regimen,
09:57typically with hypomethylating agents,
09:59such as a deciding that does
10:02show great effect in older AML
10:04patients though durable remission.
10:06Is still difficult to obtain.
10:09So going back to our case,
10:11the problem is whether we can
10:13diagnose AML in this particular
10:15biopsy where you have some areas
10:17of maturation and some not,
10:19and a absent aspirate,
10:22not just bad aspirate and just to show you
10:26in cases where you do get a good asper smear.
10:29This is the gold standard.
10:30The gold standard is counting of 500
10:32cells in each patient to enumerate
10:35not just myeloblasts and monoblast,
10:37but in the case of monocytic leukemias.
10:39Pro monocytes,
10:40which are considered blast equivalents
10:43but not to include mature monocytes.
10:46So in this study in which they had 14
10:50hematopathologist do consensus counting
10:51of a number of cases on blood and aspirate.
10:55These two are mono blasts so clearly blasts.
10:59This one is the pro monocyte.
11:02So just a step towards maturation,
11:05but still a blast equivalent.
11:07And here is a mature monocyte.
11:10And of course they use the
11:12best picture presentation.
11:13This is in a peripheral blood and
11:15just to make the situation worse,
11:18CML is further stratified.
11:20So this is the chronic counterpart to AML
11:24is stratified further by this last county.
11:29And concordance can be very difficult
11:31to achieve even among these expert
11:34hematopathologist only getting 2
11:36consensus at 74% which just to be clear,
11:40I don't think is good concordance.
11:43When you're trying to get a
11:45patient into chemo for AML.
11:47So the challenge is that even
11:49in the best possible scenario,
11:51like an excellent aspirant,
11:52the counting of Mono City lost equivalents
11:55is riddled with reliability issues,
11:58good as spirits are harder
11:59and harder to come by.
12:01This is something that when I
12:03talk to senior hematologist they
12:06really feel in their bones that
12:08it has become like a lost art,
12:10very difficult to get good aspirates.
12:12I'm on service this week and Cohen and I
12:15were guessing that about three of our 20.
12:17Last day or so we're good aspirates.
12:22Some of it is due to treatment.
12:24The newer treatments can lead to fibrosis.
12:26There are procedural issues.
12:29There's going towards IR which may not be.
12:32You know,
12:33as invested in looking at the aspirates
12:36later and there is no reliable
12:38monoblast marker on biopsy material.
12:40So just to keep in mind that aspirate
12:43where that material comes from is
12:45the same tube that the flow and.
12:48Heterogenetic S are deriving
12:49their specimen from so it kind of
12:52hurts us on multiple levels,
12:53but the core biopsy remains good.
12:55The core biopsy is still coming from
12:57that same jump shooting needle and
12:59it can add additional information
13:02beyond the aspirate in terms of
13:05architecture and localization.
13:07This is what I like to call
13:09the tree of myeloid life.
13:10So we start here with the
13:13hematopoietic stem cell going towards
13:15myeloid progenitors that then go
13:17toward GMP becoming granulocytes
13:20or monos and dendritic cells.
13:24So just in case you're wondering
13:26about mono markers in general,
13:29we do have a lot of sorry.
13:31A lot of immunostains and flow
13:34markers like these that will mark.
13:37All of these mono lineage cells,
13:42but they do not differentiate
13:45between blast versus mature.
13:47On the other hand,
13:48CD 34 is a great marker for glass
13:51that are positive for it,
13:53so the granular acidic glass are positive
13:55for CD34 whereas the granulocytes are not.
13:59So what we're looking for is something here.
14:05OK. Is there something with the sound
14:09ohh you're good now,
14:11OK? So we went on the hunt for.
14:15A phenotype. That is strongly
14:18strongly expressed by mono precursors,
14:21but not expressed in later stages,
14:24and in this older study using human
14:27umbilical cord blood and bone marrow,
14:29they were able to fractionate
14:32GMP into four subpopulations.
14:35Human common monocyte progenitors are one
14:38of the subpopulations here that do not
14:42show any potential for differentiating
14:44into myeloid or lymphoid cells,
14:46and according to.
14:48Gene expression profiling I of eight
14:50here seems to show the features that
14:53we're looking for in terms of being
14:56expressed in early mono progenitors,
14:58but not in their later stages.
15:01We also considered an R481,
15:04but didn't have as much supporting
15:07data in the literature.
15:09Fire Eight was also a great
15:10candidate for us because there was
15:13a commercially available antibody
15:15for purchase and testing.
15:16Some of you might say to yourself,
15:18wait, I just heard about IRV for some
15:22reason and you did last week when
15:25we had Lee Grimes from Cincinnati.
15:27Come and talk to us about his work
15:30in severe congenital neutropenia.
15:31He actually used IRF 8 in his recent
15:35studies as a negative control in
15:38the balance between granulocyte
15:40and monocyte differentiation.
15:42If I that paper actually came out later
15:44than when we proceeded down this pathway.
15:47But if we had seen it at that time,
15:49I think it would have given us further
15:51support to pursue this marker.
15:55So RV is a master transcriptional
15:58regulator of monocyte development,
15:59and it regulates monocyte differentiation
16:01genes that we know about is strongly
16:04induced by interferon gamma in
16:06the setting of infection and its
16:08first expressed after that comment.
16:10Granulocyte monocyte progenitor stage.
16:12Its expression is maintained at much lower
16:16levels in monos Max and dendritic cells,
16:18but not in neutrophils.
16:20It promotes apoptosis via activation
16:23of facts and repression of.
16:25CL2 and ECL Excel and loss of IRA in mice
16:29leads to an expansion of granulocytes,
16:32decreased monos,
16:33decreased's and a CML like picture and
16:37in fact overexpression of IR of eight
16:39inhibits BCR 8 ball driven leukemogenesis.
16:42It's transcripts are greatly reduced
16:44and CML patients and it's so far acts
16:48as a tumor suppressor in mouse a PML.
16:50So you might wonder what prompted me
16:53to look at this marker in an acute.
16:56Leukemia and I have to say at that time.
17:00I just really want to test it out.
17:01Given the earlier gene expression data,
17:05even though there was not much known
17:08about it acting as an oncogene in AML,
17:11and in fact it was the reverse,
17:13but later studies did show that
17:15it was a good hunch.
17:18So we started doing this validation
17:21on a cornucopia of tissues.
17:24This is susmita adapala who actually
17:26had done Hurricane Path Fellowship
17:28before she came to Yale to be a
17:31research fellow with us for a year,
17:33and she is now a hematopathologist at LJ,
17:36so she did.
17:38The scouring of literature for this
17:40gene expression profile and helped
17:42me get started on this validation.
17:45You can see that our of eight.
17:47Does in fact stain B cells in
17:50follicles in their mantle zone
17:51and in the germinal center.
17:53This is myeloid sarcoma.
17:55In soft tissue it stains the tumor cells
18:00and it is negative in this carcinoma,
18:04but you can see in the background
18:06stroma here that there are granular
18:08sites and there are histiocytes and
18:11those did not stand for our marker.
18:13So then I went ahead and pulled some of
18:16our decalcified core bone marrow because
18:19we want to use this on decalcified tissue.
18:23This is what we get most often
18:25and here the blast counting is by
18:28the aspirate or gold standard.
18:30So here is an initial diagnosis.
18:32Am OL monocytic leukemia at
18:34more than 90% blast.
18:36Here is a normal staging bone
18:38marrow for Hodgkin lymphoma that
18:40was not involved and a residual.
18:43Disease or residual for AML at 10%
18:47loss in the aspirate and here is
18:491 at morphologic remission defined
18:51by less than 5% less.
18:54So then we pulled our whole
18:57cohort of 90 am OL.
18:59That included remission residual somewhere
19:01in between and also a smaller cohort
19:06of chronic myelomonocytic leukemia.
19:08Other AML.
19:10So ammo's not monocytic and normal for the
19:15normal control bone marrows we enriched.
19:18For the ones that had monocytosis in
19:21the peripheral blood from 10 to 30%.
19:24And you can see that. This.
19:30The different diagnosis they
19:31actually had the CBC,
19:33a presentation, treatment protocols
19:35and outcomes that were compatible
19:37with what you would expect to
19:39find for each disease category.
19:44The NGS showed that the tumors
19:47had typical molecular features,
19:49about half the leukemias had NPM 1 mutations
19:53and close to a third had FLIT 3 ITD.
19:56You can also see that for
19:59the monocytic leukemias,
20:00whether they're chronic or acute,
20:03that they were enriched for SRSF,
20:052 pathogenic variants,
20:06and of course, tattoo.
20:11Two practicing hematopathologist counted
20:13the stain on core biopsies and these
20:16are plotted here for each biopsy with
20:19correlation to their aspirate blast counts.
20:22Sam Katz really gets all the credit
20:24here for prompting me to do this
20:27project in the 1st place because
20:28if you don't know him well by now,
20:31he is a very eloquent complainer.
20:34And so while some people might just say,
20:37oh, I don't like to count 500 cells
20:39or this aspirin is really bad.
20:41He really hones down on the problem here
20:45and compelled me to go upon this search.
20:50So as you might know about what they say
20:53about good deeds not going unpunished,
20:56he had to be roped into the
20:58validation as well.
20:59So this was done independently
21:01with disregard for their diagnosis,
21:04and we achieved a pretty good correlation.
21:08This was the diagnostic test
21:11characteristics using aspera count
21:13as the surrogate for disease status,
21:16so AML being 20% plus or higher
21:19residual disease being 5% plus or
21:22higher and negative or residual being
21:25less than 5% as compared to IR 8 IHC
21:30result due to a reviewer question.
21:33We actually went back and did the
21:35same with CD 34 because we actually
21:38didn't know how well we're doing.
21:40CD34,
21:40as opposed to our aspirate blast
21:42count in the granulocytic leukemia
21:44and we did not actually get to
21:47the same good correlation.
21:48It was still good,
21:49but it was not quite at .8
21:51which we had for IR 8.
21:56And this is the correlation for CML,
21:59which is not as strong.
22:00But we also had a smaller cohort for CML.
22:04One of the reasons that CML cases might
22:08be especially difficult I believe,
22:11is that occasionally,
22:12as with our first case that I showed there
22:15is focal elevation of glass which may
22:18not be well represented on aspirin smear,
22:21because, as you might know,
22:24for aspirus you're really kind of sucking it.
22:26Thought from one specific point.
22:28So here is an interesting biopsy.
22:30We had a few years ago where you can
22:32see even on low power that this corner
22:35here looks different from this part.
22:37So most of the bone marrow showed
22:40as in the upper part maturing
22:42trilineage amount of polices.
22:44Some increase in boss because
22:46this person also had CML.
22:48That was a little bit elevated
22:50like CML one or CM L2,
22:52but then the lower right hand corner
22:54actually had a sheet of glass.
22:56As represented in E here and when
22:59we did our of eight you can see
23:01how dramatic this transition is.
23:03Almost like a solid tumor malignancy.
23:09I also get asked whether the
23:12macrophages are positive,
23:13and I think not.
23:15This is a AML that is not
23:18monocytic with a federal flage.
23:21Here this is a monocytic leukemia
23:23at initial diagnosis and this is
23:25one of our staging bone marrow
23:27biopsies showing a Cidra page here
23:29and they were negative for a marker.
23:33There were discrepancies in
23:35assessing residual disease,
23:37so 10 cases showed a discrepancy defined
23:41at where aspirate had less than 5% loss.
23:44But IRA expression was
23:46just a little above 5%,
23:49so three of these actually had
23:51definitive evidence of disease by
23:54cytogenetics as unbalanced translocations
23:56flit 3 ITD or MPM 1 mutations,
24:00and three have clinical relapse
24:01within two months of the biopsy.
24:04So I just want to point out here
24:06that our whole correlation our
24:07ground truth in this study has
24:09been the aspirate blast count.
24:11But that as we know,
24:13as pathologist is just another sample.
24:15It's a different sample from the core biopsy.
24:17So what is really the truth?
24:20Maybe it is the clinical behavior.
24:22Maybe it is genetics or molecular and
24:24and I think that that correlation comes
24:27into play when we're looking at a new marker.
24:30One had flow elevation of
24:32human tacones above 10%.
24:33Which pumped in me to go on a search for
24:35all the biopsies that had increased humidity.
24:38Phones, hard to find,
24:40but they're they are.
24:41IRA dusting hematogenous,
24:42so that could be a pitfall.
24:45In the rare case where you had
24:48really significant elevation by
24:50flow and then we had two remaining
24:52discrepancies of overcount by IR.
24:54RF eight as compared to aspirate
24:56at a low percentage blast,
24:59one did show RA expression that was
25:01lower than the aspire blast count.
25:03But upon evaluation,
25:05the core biopsy showed
25:07substantial aspiration artifact,
25:09so perhaps it should have been
25:11excluded in the first place.
25:12So, excluding the above cases,
25:14where disease was still present,
25:15the overall discrepancy was 3% with
25:19regard to residual disease assessment.
25:22So in conclusion,
25:23in Almal there is high correlation of IRV
25:26positive cells to aspirate last count.
25:29The comparison of IRV staining to
25:32blast count and see mammal also
25:33showed a pretty good correlation,
25:35though that requires more study and
25:38in contrast reactive monocytosis and
25:41AML without monocytic differentiation
25:43did not show IRA elevation when it was
25:47used to categorize cases as acute leukemia,
25:49positive or residual leukemia or negative.
25:52Sensitivity and specificity
25:53was high and this marker can be
25:56clinically useful as IHC to possibly
25:59diagnose and track disease.
26:01Particularly in cases of poor
26:04aspiration or focal blast increase.
26:07So shortly after this manuscript
26:09was accepted for publication,
26:11actually a great study came out
26:13in molecular cell that actually
26:15really helped me understand
26:16what we were seeing better.
26:18This is a summary of transcription
26:20factor domain focus.
26:21CRISPER screens genes were ranked
26:24by AML biased essentiality scores
26:26defined by the difference in
26:29a particular domain score in
26:31AML versus non AML cell lines.
26:33And I should point out that
26:35in the cell lines
26:36that showed. Particularly high IRF
26:398 dependency, those were monocytic,
26:42leukemias, and the others are not.
26:48Umm? Competition based proliferation
26:51assays performed in Castine positive
26:54cell lines show that AML cells
26:57that were transduced with IRF 8
26:59guided RNA were rapidly depleted
27:02and outcompeted by parental cells.
27:05In separate studies showing
27:09173 AML patient samples,
27:10High R of eight expression was seen in
27:14association with diverse cytogenetic
27:16and molecular driver mutations,
27:18so there's genetic and molecular features
27:20actually did not define this behavior.
27:28They also found that RFA is
27:29enriched at the MEF 2D locus.
27:31That is another transcription factor,
27:34and it is known to be important in
27:38some Bal in several cell lines,
27:41such as month 13,
27:42which is a monocytic cell line,
27:44RNA seek of gene expression.
27:45Changes indicate that cell
27:47lines transduced with guided
27:49RNA to IF8 revealed MEF 2D to
27:52be significantly downregulated.
27:56So transcription factors are
27:59difficult to therapeutically target,
28:01so they looked for a chromatin
28:03regulator upstream that would be
28:05more druggable in a separate crisper
28:07dropout screen that was focused
28:08on chromatin regulatory domains.
28:10They uncover CMU and eight as AML specific,
28:14and here you can see that in certain
28:17cell lines they are hypersensitive to
28:20depletion of this chromatin reader,
28:22and others are less so.
28:25The ones that are non AML are
28:27insensitive to it and by the way,
28:30cmda is ubiquitously expressed
28:31in all cell lines,
28:33so it was not because of how
28:35much the YD ate there was.
28:39They further found that all the
28:41AML cell lines are IR 8 high are
28:44hypersensitive to depletion of CMY D8
28:47and loss of CMD 8 resulted in decrease
28:50in ire of eight as well as Mick
28:52over time and an increase in Miller
28:55differentiation associated genes.
28:57Genetic depletion CMND in normal
28:59hematopoietic stem cells did
29:00not impact cells in vitro.
29:02So just what you want for a druggable target.
29:10Altogether and with other data that I'm
29:11not presenting in the interest of time,
29:13they show that IR 8 helps form
29:16a transcriptional circuit to
29:18support AML proliferation.
29:19They also showed that a targetable chromatin
29:23reader CMI and eight regulates IRF 8
29:26by lineage specific enhancers in AML.
29:29So at this point we were quite
29:31convinced that I of eight might
29:33serve as a transcriptional addiction
29:35in monocytic leukemias,
29:36and we were curious as to how else we
29:39can use this in daily clinical practice.
29:42What I'm presenting here is work
29:44done by Dan Mcquade,
29:46who is the 2nd year Yale MD,
29:48PhD candidate who really learned
29:51all of heme path.
29:53It seems within a year and this was
29:57just published a couple months ago.
30:00We showed that RF-8 specifically stains
30:03moneyglass and these extramedullary tumors,
30:05so this is a myeloid sarcoma in a lymph node.
30:08Follicles are still intact,
30:09but you just can see the blast.
30:12In the uh, Paracortical area,
30:14some of the blocks are more towards medullary
30:17area and they were actually MPO positive.
30:20Seen here,
30:21but the areas of the boss without MPO
30:25positivity were expressing IRF 8,
30:27so this is in tumor.
30:29This is a little dimmer in the
30:31mantle zone and when we plotted the
30:35relative expression of CD 34 MPO
30:37and RF-8 for all of these cases,
30:40you see an almost inverse relationship.
30:45Another recent I think that we did was
30:50looking at BPDCN, so BPDCN is another
30:54extramedullary hematopoietic tumor.
30:56It's often difficult to diagnose
30:58as it presents first in the
31:01skin or soft tissue and the most
31:04helpful marker to date is CD 123,
31:07which is also a druggable target.
31:09This can be very dimly staining.
31:11This is not just in our lab,
31:12it's universally known that city 123
31:16can be somewhat dim in this tumor.
31:18That really, really relies on the marker,
31:21so we look to see if IRF 8 can be
31:23helpful in these instances and it was.
31:25This was done in collaboration with Doctor
31:28Gallery Pansy from Dermatopathology here.
31:34Doctor Jacqueline Pinkus,
31:35who was the health director at
31:38Brigham when I was training.
31:40She is just a guru in immunohistochemistry.
31:44I think she actually discovered
31:45how disdain for Kappa,
31:47Lambda and tissue back in the
31:49day and when she saw our paper,
31:51she immediately started doing some
31:53double stains and this is leukemia cutis.
31:56Double stain with RV in
31:58brown and lysozyme in red.
31:59So as you might remember,
32:00lysozyme stains all of the different
32:04maturation stages of monos and histiocytes.
32:07And you can see it staining
32:09the history of site.
32:10Like Sweets,
32:11but there are eight is not staining
32:14because there are no glass in here.
32:16Other potentially helpful double
32:18stains are CD34 with RF-8.
32:21I think in terms of what if we
32:24have a Milo monocytic leukemia and
32:27I think Doctor Pinkus is already
32:30doing the double stain with City
32:32123 to kind of take out the
32:34dendritic cells in a bone marrow.
32:38We also looked at the TCG,
32:40a pattern of expression for IRF 8.
32:43The red block boxes are tumor.
32:45The Gray is normal, and you can see
32:48for AML and DLBCL you have this very
32:51dramatic difference in expression,
32:53whereas in the other cancers,
32:55not as much in the ones
32:57here like this is, I think,
32:59stomach and testicular germ cell tumor.
33:01We're kind of curious as to whether
33:03it really did show that difference.
33:05We had a tissue microarray composed of.
33:08All these different types of carcinomas
33:10and it really did not stain for any of
33:13them except for one lymphoma that snuck in.
33:15This was actually diagnosed by Doctor
33:17Jose Costa really back in the day
33:20where it was called malignant lymphoma.
33:22So of course that prompted me to do a CD
33:2520 just to make sure it was in fact a DLBCL.
33:28So for the rest of the tumors you can
33:29see here that there is some staining,
33:31but that is not in tumor,
33:33so that may explain the TCGA data.
33:38We also looked at the other
33:40differential diagnosis in soft tissue,
33:41which is actual sarcomas as
33:44opposed to myeloid sarcomas,
33:46so this is done in collaboration
33:48with Doctor William Wong,
33:49who had been a mentor of Doctor Gary
33:52Pansies when she was in training,
33:54and you can see that for the sarcomas,
33:56including the ones that are
33:57small round blue cell tumors.
33:59They're negative for higher of eight.
34:02So so far we've been talking about all
34:04the myeloid and monocytic differentials.
34:07What about the lymphoid?
34:08Because we know it also stains for B cells.
34:12Well, here is another common
34:14problem in hematopathology classic
34:15country and lymphoma versus
34:17anaplastic large cell lymphoma,
34:19and these have overlapping
34:21morphologic features.
34:23They also have overlapping
34:25immunophenotypic features,
34:26and we sometimes have to rely only on PAX 5,
34:30which is not always.
34:32Positive and classic caution lymphomas
34:34and can also be amplified in some alcl's.
34:36So we took 74 cases of Hodgkin and 15 cases
34:39of ALK negative LCL to see how they would do.
34:42Obviously all positive ACL we
34:44would not have a problem with.
34:47And you can see here that even in the
34:49PAX five negative Hodgkin lymphomas,
34:51you can have dim staining.
34:52For IRF 8,
34:53we shouldn't be using this in isolation,
34:55of course,
34:56but in the context of the
34:58morphology and other markers,
35:00I think it can be helpful since
35:02it is dead negative in all ACL.
35:05And this is the kind of
35:08composite of cases are pacified,
35:10negative I of eight positive,
35:12some of course are double negative,
35:14and here we really have to look
35:16at all the different stains and
35:18other features that we have.
35:20This was also a first author paper
35:23for Dan who just got this accepted
35:25for publication couple weeks ago.
35:27So in conclusion,
35:29IFA can be used to detect extramedullary
35:31hematopoietic tumors as well.
35:33That can be a diagnostic challenge
35:36such as leukemia cutis myeloid sarcomas
35:39BPDCN its expression is essentially
35:41absent in all other solid tumor
35:43malignancies that can present as a
35:45differential diagnosis here and as
35:47a transcription factor that's also
35:49important in B cell lineage commitment.
35:51It can show some promise in the
35:54detection between Hodgkin versus ALCL,
35:56which is a real clinical dilemma.
36:01Our findings were just recently
36:03replicated by a couple labs in
36:05Switzerland and Italy, showing that yes,
36:08it is a reliable monoblast marker,
36:10but they also show it staining BPDCN well.
36:16And thanks to Anoj Verma, our resident here,
36:19this has been reviewed in the context
36:22of other emerging immunohistochemical
36:25biomarkers for myeloid neoplasms.
36:30So what about our case?
36:31Or a patient with CML?
36:33How do I diagnose this as our
36:35final diagnosis after showing this
36:37case around to multiple other heme
36:39path faculty members was myeloid
36:41neoplasm with increased blasts most
36:43compatible with progression to AML?
36:45And our patient was started on treatment
36:48with PIXIUS which is a liposomal 7 + 3.
36:52It's Donna Robinson was cytarabine
36:54with a consideration toward
36:56associated Gene and venetoclax.
36:58She would have gotten that instead if.
37:01The age was older and she
37:02didn't have the symptoms.
37:05So here is our case.
37:07When I went back to stain it with our eight,
37:10you can see that in fact the
37:11areas where we suspected increased
37:13loss there was increase in RF-8
37:15and in the areas of maturing
37:18Trilineage Marquis there was not,
37:20so this certainly made me feel better
37:22about calling her as evolution to AML.
37:26Another recent study also showed I of a
37:29as an AML specific susceptibility gene.
37:32So here using publicly available
37:35databases they show that these red dots
37:39representing genes that were essential
37:41to AML but not in non AML cell lines.
37:45And here what I really want to
37:48highlight is that IFA expression is
37:51also correlated with poor prognosis and
37:53I think for those of us who remember.
37:56The FAB classification and little
37:58period of time after that,
38:00many times when we first diagnosed AML,
38:03clinicians will come down and say,
38:05but is it monocytic?
38:06Does it look monocytic to you?
38:08Because then I will consider
38:10this as a worse prognosis.
38:12And here we have if we consider
38:14RFA to be a marker of mono bus,
38:16some substantiation of that suspicion.
38:19We're no longer asked that because
38:21there's more of a focus right
38:23now on exactly what mutations
38:25and genetic translocations?
38:26They have,
38:27but I think that that older clinician
38:30perspective of when we say monocytic,
38:32it's usually worse for them.
38:34Still holds true in some way.
38:38This also made me think of a prior
38:41observation made a couple years ago,
38:43which I think is experientially
38:45many hematopathologist feel that we
38:47have seen certain very prominent
38:50instances of which is monoblast elk
38:52growth after venetoclax treatment.
38:55So in this study they also segregated
38:57their patients based on fab
38:59classification and found that only the
39:02monocytic classifier was predictive
39:04of refractory response to Aza Ven.
39:06So here.
39:07This is a non monocytic leukemia
39:10and this is a monocytic leukemia.
39:13They talk primary human AML cells from
39:16both what they term primitive and
39:19AMOLED and in vitro demonstrated lower
39:22kill with both phonetic clocks and a
39:26combo phonetic clocks with acidity.
39:29They also showed this in a violin
39:32plot how monocytic disease arising
39:34after treatment can be derived
39:37from preexisting subclones.
39:38In this other patient at relapse.
39:43So our future directions include
39:45something Poe hands working on with
39:48flow cytometric detection of R8
39:50and refining that monoblast gate.
39:53Doing that in collaboration with OHSU,
39:56a multi institutional validation
39:57using AI tools that suit your parent.
40:00Cherry is leading and I got a lot of
40:03helpful consultation from Doctor David Rim
40:06and this is being done in collaboration
40:09with MGH BWH Upenn, New Mexico.
40:13OHSU Cornell and Stanford and there
40:15is a focus in this group on wanting
40:18to further look at CML and this
40:21whole subclassification of CML which
40:24is a bit controversial in The Who.
40:27Whether this marker can help
40:29us hone it down a little more.
40:31And I do hope to identify IRF 8
40:34target genes in actual am OL samples
40:37and in primary human monoblast.
40:39And since I have a regulates BCL,
40:41two family members maybe 1 pathway.
40:44Of venetoclax resistance in AML.
40:48So thank you very much for
40:50your attention and your time.
40:53Thank you so much for the many people
40:55who contributed and really helped
40:58move this forward to our external
41:01collaborators for the project that
41:03our that is ongoing and to our
41:06colleagues in derm path soft tissue
41:09path pathology tissue services this
41:11could not have been done without Amos
41:15and Laurie and are flow hematology.
41:18And biostats colleagues.
41:20Thank you so much.
41:30We are open for questions that
41:33was excellent by the way.
41:36From a non humanoid pathologist,
41:39I thought the story is riveting.
41:43Thank you.
41:45May I ask
41:46a question Mina that was fantastic,
41:48lot of work and you know you have carried
41:51through something that you found and
41:53it is now evolving into something far
41:55bigger than you may have imagined.
41:57Initially from the stain that you
42:01showed you in the recent case of the
42:05case that you're diagnosed with acute
42:08or progression to acute leukemia.
42:12It looks like the staining.
42:16Is was probably more than 20% and
42:18there is a graduation of stating
42:20there are many nuclei which are
42:22dimmer in their expression of ifit so.
42:27Do you think as the last mature,
42:30the expression level slowly decreases?
42:32Is not an all in non phenomena and
42:35second question is this is a surface
42:37you know marker in some ways right so?
42:42It's a it's a nuclear marker
42:43nuclear marker, so is it? So can
42:46it be targeted for therapies
42:47in some ways or no?
42:51OK, so great questions Dan Pat.
42:54I think for the first one
42:56you know in terms of the.
42:58Some lighter or some darker that is
43:02very important and I think that.
43:05In the validation, what was helpful
43:07was its correlation with last count.
43:09So we were counting all the ones
43:12that had any staining and you know
43:15obviously this is going to be.
43:20You know, if you.
43:21Had a stronger dilution you might
43:23be seeing more of the background
43:25if you had a lesser dilution.
43:27So what we titrated to was that point
43:31at which we were seeing a percentage
43:35blast that were reflective of the
43:38actual morphologic blast count.
43:40So it is interesting in that a
43:44lot of hematopathologist ask,
43:46well, does it stain Pro monocytes,
43:48and it's such a good question.
43:51I don't know how to answer it because.
43:53Homicide is a cytologic definition
43:55when we see it in the aspirate
43:58and there is no great marker that
44:01just gets at the Pomona sites.
44:03So I wish that I could just stay
44:05in and aspirate that had a ton of
44:07promyelocytes and see if it did pick them up.
44:10The only consoling part of this
44:13whole dimmer staining is that what
44:16we got it to is in correlation
44:18with that gold standard,
44:20so we are counting them and I do think that.
44:23At some point it is not
44:26seeable by light microscopy,
44:29but is probably still baseline there by RNA,
44:34right?
44:34Like if you look at the really
44:36mature macrophage you can see in
44:38the gene expression profiles that
44:39they do have some expression,
44:41but we're not seeing it by by our IHC now.
44:45That particular case, if you use that state,
44:48your counts for the blast would have
44:50been much higher than 30%. Then
44:52is that right? Or I'm just?
44:54Yeah, I think what I showed
44:56you though was one foci right?
44:58One area where it seemed a little higher,
45:00but then just in practice,
45:03and I'm not sure this is the
45:04right thing to do in practice.
45:06By consensus we we typically have to
45:09do it through the whole entire core.
45:11Yeah, and in terms of whether
45:13it's therapeutically targetable,
45:14you know it may be,
45:15but it is a transcription factor and it is,
45:18you know, also important in
45:20B cell lineage development,
45:21so I think it would be hard
45:23to just try to focus on that.
45:26You could end up causing a lot
45:29of the problems.
45:30David, I think you were next with the hand.
45:33Yeah, that my question is similar to
45:35Dan Potts in a little bit in a little
45:37ways I see that you're progressing
45:38toward a multi institutional study
45:40which will be really interesting to
45:42see if this can be carried out by
45:44many pathologists at many places,
45:46but it worries me that the intensity
45:48at which excel becomes a positive
45:50cell because someone were pretty
45:52light and some of them are pretty
45:55dark and that probably represents a
45:57difference in RF-8 expression levels.
45:59Do you have any way to standardize that?
46:00Or how are you going to check that the
46:02results here at Yale are the same?
46:04As your other institutions that
46:05you're collaborating with,
46:07yeah, that's that's a great question,
46:09and we are. You know, as you know,
46:13we're trying to do this with some
46:15quantitative imaging analysis to see out
46:18what cutoff point does it actually show.
46:22Kind of a consensus with regard
46:25to not just the diagnosis,
46:28but a consensus across all 7 institutions.
46:33And you know, I,
46:36I think that this is I wish we
46:38had done this earlier with CD 34.
46:40In fact, because all these labs.
46:42Doing CD 34 and using that day-to-day
46:44to tell you whether the AML still there
46:48but we don't know how compatible we are
46:51with each other and whether you know
46:53if we did some quantitative imaging
46:56whether it would actually bring us
46:58to a better consensus and tell us the
47:00truth that we're not reporting right now.
47:03So I and you know,
47:05nobody really wants to do the bus
47:07quantification themselves by I,
47:09so I think it's it's really moving us
47:11forward to a point where we can be.
47:13More standardized across the board.
47:15Yeah, I'm not worried about
47:16standardized accounting as much as I am
47:18standardizing the biochemistry part,
47:20that is the tighter and the stain.
47:22That is, how do you know
47:23if it's if it's too light,
47:25then whether you count it by
47:26eye or by an imaging system,
47:29it will just not be counted,
47:31and that's what I would guess that there
47:33will be some institutions that will be
47:35lighter than other institutions overall,
47:37and then that could throw off the result.
47:40Absolutely yeah.
47:42Transit control,
47:42like you know with estrogen
47:44receptor we have intrinsic the
47:46ducts inside the normal breast
47:48and we absolutely have
47:50those internal controls.
47:51That's the good thing with him is
47:53we almost always have internal
47:54control because there's so many
47:56other cells in the background.
47:57So we're using B cells as the
48:00control and also dendritic cells.
48:02So I've actually seen
48:05brighams immunostain for it,
48:07which you're using on the Ventana I believe.
48:11And also cornells and also.
48:16OHSU UM and and then of course the
48:20two international groups so so far,
48:23seven groups in the US large
48:26academic centers have brought it on
48:29board for optimization and I have
48:31not yet seen one that was very,
48:34you know, significantly different
48:36in terms of staining a tissue.
48:38They've also sent me their tissues
48:41to stain to compare side by
48:43side so that has been helpful.
48:45It takes a lot of work to see.
48:47Whether that question and
48:48that answer holds out,
48:49though it's a great question.
48:52Amarie
48:54thank you, just a superb talk and
48:57such an elegant presentation of your
49:00thought process and the steps through.
49:03And so I congratulate you as a
49:07busy surgical pathologist for
49:09thinking for the thought process
49:11and for getting this work done.
49:13So many congratulations.
49:14I I'm I think a lot about inflammation
49:17these days and I notice that I
49:20don't know anything about RF 8
49:22except I do notice that it is.
49:25Important in battling infection
49:26and that if you if you're deficient
49:29in IRF 8 you have a severe primary
49:32immunodeficiency and that it's
49:35also now in GW studies,
49:38they find variance of IRA or a
49:41significant risk for autoimmune diseases,
49:44and I just wondered if you had
49:46any any thoughts about that?
49:48Does that manifest at all in your
49:50world and pathology and and I'm?
49:52I'm wondering if we can use
49:54this antibody as well.
49:55As we think about these autoimmune
49:58inflammatory conditions.
50:01Yeah, I that is such a fantastic question.
50:04I really don't know that much about its
50:08role in autoimmunity except to say that you
50:12know if you take out the monocytic lineage,
50:16there will be kind of hell to pay, you know.
50:19So I think that it it absolutely plays
50:21an important role when it's functional.
50:24When it's doing its normal job in
50:27terms of infection and inflammation.
50:29So we cannot.
50:31Completely take it out of function,
50:34but only in when it is upregulated in tumor.
50:39And I think that's it.
50:41Also is confusing to me and I think something
50:44that I really want to work on is why in in
50:47its normal function it is proapoptotic.
50:49But then in tumor,
50:51obviously they are continuing to
50:53proliferate and live on, so it must be.
50:57Maybe it's binding differently in some way.
51:00When it's tumor,
51:02you know there there have been
51:05studies showing its deficiency and
51:07DLBCL that I have to read more about,
51:09but you know it has manifested
51:12different kinds of phenotype,
51:14whether it's tumor versus normal.
51:16So I hope that others will be
51:18interested in working on this because
51:21I cannot do all of the different
51:23types of studies and IRA.
51:25I think it's just really cool and.
51:27Complicated.
51:30Won't you?
51:32I've got two questions. One is,
51:35do you think that bone marrow aspirates
51:38are going to fall out of favor?
51:45So, so I don't think so.
51:48I you know it's hard for them to do
51:51a good aspirate, but a good aspirate
51:53is just worth its weight in gold.
51:55It is so important to get that
51:58fluid sample that we can use for
52:01cytogenetics and molecular and flow.
52:03I don't think it would ever
52:04really fall out of favor.
52:06I do think that.
52:07We need a better way to obtain it
52:09because it is not just like at Yale or
52:12you know this clinic or that clinic.
52:15When I've talked to other institutions,
52:18this is a major problem in the US and
52:21folks coming from outside of the US
52:25hematologist trained at other countries
52:27usually have been really trained at
52:30looking at their own aspirates and
52:32seeing whether they're good or not.
52:34They're much better at
52:36getting an aspirate right,
52:37so I think it's partly that
52:39the training and the US.
52:40Focus more on treatment versus
52:44the diagnostic procedure,
52:46so there might be a swing in
52:48the other direction as we see
52:50how important the aspirate is.
52:53The other question I have is
52:56this drug that works again.
52:59That targets BCL.
53:01Two BCL two is expressed on
53:04other lymphocytes as well.
53:06So are there what happens to
53:10those lymphocytes in patients
53:12that are receiving this drug?
53:16Yeah, you know no drug is
53:18without its side effects,
53:20but I I have to say that venetoclax
53:22has reportedly done pretty well in
53:25these patients because so much of
53:28their lymphocytes are actually really
53:31abnormal or myeloid cells are so
53:34abundant that they have a depression
53:37in normal lymphocytes already.
53:39So venetoclax has been shown to
53:43really induce that initial remission.
53:46The duration of that remission is not long,
53:49so that is a problem.
53:50And these breakthroughs and relapses
53:54happen pretty quickly thereafter.
53:56So so that is a real issue.
53:58But, you know, I think also we,
54:02we think of OK, this stain is high,
54:05and so this this drug must work, you know.
54:08Well in the one.
54:09Sustained highly,
54:10but it actually has not worked out that way.
54:13So in terms of CLL it's worked really well,
54:18but you know,
54:19full of your lymphoma is BCL 2 positive
54:21and yet it's not as it's not doing
54:24as well in follicular so it's not
54:26like a one to one correlation either.
54:31Thank you Mina. The venetoclax
54:34is very interesting.
54:35Manju, because this is a drug where you
54:38actually have to titrate up in CLL,
54:39you give it all at once and patients
54:42will get extreme tumor lysis syndrome.
54:44The problem with these drugs as well,
54:46based on their structure
54:49is they're extravascular.
54:51Extrusion is very limited,
54:52so there are many tumors where within the
54:55circulating system it's very effective,
54:57but on tissues it's just a drug property.
54:59That drug is very,
55:01very tightly albumin bound and
55:02very poorly able to actually
55:04egress from the bloodstream.
55:07Thank you, thank
55:08you so much. That's a great point.
55:11Hi Mina, I have a question for you.
55:15So this is more of an
55:18immunohistochemistry question.
55:19Are you really looking into multiplexing?
55:22And since this monoblast count
55:24is so critical in the 20% count,
55:27so I was wondering,
55:28are you relying only on IR F8 and would
55:32you also consider multiplexing with CD 34?
55:35But then we no CD 34 doesn't pick
55:38up all monoblast monocytic leukemia,
55:40so is there any other marker for example,
55:44which is cytoplasmic localization?
55:47Like CD 117 would be more helpful
55:49for you versus CD 34, which also I
55:52believe goes to the nucleus, right?
55:54So having a Multiplex stay in that is.
55:57You know,
55:58picks up two different localizations
55:59would be more helpful than having.
56:01Absolutely yeah,
56:03I'm I'm so glad you brought it up
56:05and I really hope to get your buy in
56:07and doing the multiplexing. In fact,
56:10I think CD34 with RA would be perfect.
56:13In fact, CD 34 is going to be membranous,
56:16and this is nuclear.
56:17Then we would capture all of them, you know.
56:21And at MGB they're already doing 123 with RH,
56:26just to show its double.
56:28Expression in the dendritic cells,
56:32so so that's another possibility.
56:34But I I do think that in cases where
56:37we really are relying on this kind of
56:41getting to 20% the the multiplexing
56:43with 34 would be super super helpful.
56:46So I'd be happy to work on it
56:48with your staff.
56:53Thank you so much for all of your time.
56:57Please feel free to ask me any other
56:59questions as you might see me in
57:00the hallway. Thank you, thank you,
57:04thank you all for coming.
57:07Great talk.