Yale Pathology Grand Rounds: October 13, 2022

October 13, 2022

Information

Quest for Perfection: Updating the Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) presented by Syed Z. Ali, MBBS, MD

ID8165

To CiteDCA Citation Guide

00:00I think we'll get started.
00:02So it's 1230. Thanks all for coming.
00:04It's my great pleasure to
00:06introduce friend and mentor,
00:07doctor Syed Ali's world renowned
00:10cytopathologist and innovative educator.
00:12Doctor Ali serves as professor of
00:14pathology and radiology at the Johns
00:15Hopkins University School of Medicine as
00:17well as a director of the Division of
00:19Cytopathology at Johns Hopkins Hospital.
00:21He received his pathology residency
00:23training at North Shore University,
00:25Cornell University in New York
00:27with fellowships and oncologic
00:28surgical pathology at Memorial.
00:30In New York,
00:31under the mentorship of Doctor Juan Rosai,
00:33followed by cytopathology at
00:34the Johns Hopkins Hospital
00:36in Baltimore. Known globally
00:38as an innovative educator and one
00:39of the most traveled pathologists,
00:41doctor Ali has well over 350
00:44invited lectures, keynote speeches,
00:46and visiting professorships
00:47all over the world.
00:49In collaboration with international
00:51societies in universities, he directs,
00:53or Co directs several annual major
00:56cytopathology tutorials.
00:57He's published extensively with over 300
01:00articles and major scientific journals,
01:02over 35 books, book chapters,
01:04and electronic media.
01:05Known for his expertise in thyroid
01:07and pancreatic cytopathology.
01:09His monograph on the Bethesda system
01:11for reporting thyroid cytopathology
01:13has been translated in seven
01:15languages and is extensively used in
01:17medical practice all over the world.
01:19He's the founding editor in chief of
01:21the Journal of the American Society
01:22of Cytopathology and has served as
01:24a member of the Executive Board and
01:26chair of the Scientific Program.
01:28And other committees of the
01:30American Society of Cytopathology.
01:31He's past president of the ANC,
01:33presently serves as Vice President
01:35in November,
01:36president-elect of the International
01:38Academy of Psychology,
01:40and the Chair of the International
01:42Board of Cytopathology of the AIC.
01:44He's a fellow of both the Royal
01:46College of Pathologists and the IC,
01:47and a recipient of ASC's Warren
01:50Lang Award in 96,
01:51the President's award in 2012,
01:54Excellence in Education award in 2015,
01:56and the Maurice Goldback Board.
01:58In 2019.
02:00I personally know Doctor Lee for
02:01over 25 years since my residency
02:03training at Johns Hopkins.
02:04I can personally speak to the impact
02:06he's had on those who he's trained
02:09and cytopathology education overall.
02:11As he shared with our residents this
02:13morning at the diagnostic seminar,
02:14mentors like Doctor Rose I have
02:16a lasting impact on trainees,
02:18pathologists, and the profession.
02:19Like Doctor Rosai,
02:21Doctor Ali continues to challenge us.
02:22He not only lights up a room,
02:24but also past Twitter with fascinating cases,
02:27morphologic observation,
02:28detailed descriptions.
02:29That extend to not only us,
02:32but to over 25,000 of his
02:34international pathology followers.
02:35Without further ado,
02:37I am privileged and delighted
02:38to introduce Doctor Syed Ali.
02:40His topic today Quest for perfection,
02:42updating the Bethesda system for
02:45reporting thyroid cytopathology.
02:49Thank you so much, Angelique,
02:51for that kind introduction.
02:53Allow me to say that you were
02:56one of the best that you trained
02:58at Hopkins and it's so good to
02:59see you do so well here at Yale.
03:01So we do miss you.
03:04Good afternoon, everyone.
03:05It's a pleasure to be here,
03:07especially presenting a topic to you
03:09which is very near and dear to my heart.
03:12As Angelique mentioned,
03:14we are in the process of updating the.
03:17With this system,
03:18which we will call 2023 Bethesda System or
03:22the 3rd edition of the Bethesda system.
03:26As my friend Mag Demay often says
03:29that I don't have any conflicts
03:32of interest and I don't have
03:35any interest in conflicts.
03:37So before I proceed,
03:39I just wanted you to read what the Great
03:43Leo Post said in his very well known book,
03:46the 1st edition,
03:47and it will tell you that the reporting
03:51or the language that we use when we
03:54report the psychological findings are
03:57as important as the diagnosis itself.
04:00I often tell my own trainees that
04:03it's 50% of the job that we do as
04:06cytopathologist diagnosing something.
04:08The way we properly communicate
04:10that diagnosis to be clinician
04:13is the remaining 50%,
04:14because if you don't use the
04:16right language or terminology,
04:18your diagnosis of excellent
04:21diagnosis could be misinterpreted.
04:24It's valid in search path,
04:25but it's much more valid inside of pathology,
04:28where typically no intervening
04:30biopsy is done and the affinity
04:33diagnosis most of the time will lead
04:36to definitive treatment or surgery.
04:38I'm sure people who sign out
04:41psychology would know that the tester
04:44system was based on a lot of work
04:46which was done in 2007 at the NCI.
04:49NCI is located in the town of
04:51the Test in Maryland and has the
04:53name of the tester system.
04:55It took us three years to
04:57publish the first monograph,
04:58the 2010 edition, which is shown here.
05:01And then of course another seven years
05:04passed before we updated the system,
05:06which we call the 2017.
05:08Condition.
05:09And as you can read here,
05:11there were there was no significant
05:13change in the six terminologies
05:15or six groups that are used to
05:18report thyroid psychologies,
05:20non diagnostic or unsat.
05:22We have benign AUS or plus SFN or FNS,
05:26FM and malignant.
05:28I'm just abbreviating to save time.
05:31But you would also notice something here,
05:33which is unprecedented,
05:34that we were in a consensus meeting
05:37at NCI and we did not agree
05:39on having single names.
05:40Three of the six diagnostic categories,
05:42and the main reason was that that we
05:45wanted to be completely inclusive
05:48of everyone's concept and approach.
05:50And I build a group of us thought that
05:53non diagnostic is better than unset,
05:56or AUS is better than flask,
05:58so on and so forth.
05:59So we endorse both terms would
06:02be strongly recommended to the
06:03cytopathologist that use one or the other.
06:06Don't use the two terms
06:09interchangeably because.
06:11You will then confuse the
06:13clinicians and defeat the purpose
06:15of having a nice reporting
06:17system as Bethesda.
06:18But down the road we realized that
06:20a UAS and frost was causing most
06:22of the diagnostic confusion and
06:24I'll address that in a minute.
06:26The other thing which you will notice
06:28here is that there are no numbers written
06:30before these diagnostic categories.
06:32Bethesda System was never meant to
06:34be a numerical reporting system.
06:36It has beautiful names, brief names,
06:38easily memorizable.
06:40And and usable,
06:42and hence those of you who use numbers
06:45I would strongly urge that also puts
06:47the proper name after the number.
06:50But this number don't convey much meaning,
06:52and history has shown that
06:55numerical reporting systems often
06:57fail because of of that reason.
07:00So this is the update in 2017 with Tessa.
07:03So everything which is written
07:05and read was an update.
07:07And as you can see here,
07:09the risk of malignancies which are
07:11implicit in the Bethesda system.
07:13And that's the main reason of the
07:15success of the there's a system
07:16that was built on a framework
07:18of probabilistic approach,
07:19meaning probability of finding malignant
07:21tumor or ROM is built-in into each
07:24of the six diagnostic categories.
07:26So it was significantly revised
07:28based on prospectively.
07:29Analyze data,
07:30it went up from 1:00 to 4:00 to 5:00
07:33to 10:00 and and as you can see the
07:36other three categories also showed a
07:40significantly updated ROMs and also
07:43the management algorithm change.
07:45As you know that ATA American Thyroid
07:48Association has completely endorsed
07:50Bethesda System as as most of their
07:54recommendations management plans
07:55are based on the test a reporting
07:58system and we realized that.
08:00After 2015 when the IATA guidelines
08:03were revised,
08:04molecular testing was introduced,
08:06a more conservative approach lobectomy
08:08was introduced and again molecular
08:11testing for follicular neoplasm.
08:12And one of the recommendation 2015
08:15with Test 80A guidelines was that
08:18if the thyroid tumor or nodule is
08:214 centimeters smaller with a well
08:23differentiated thyroid cancer either
08:25follicular and papillary with no
08:28extracapsular extension or nodal. Disease.
08:30In other words, see and not disease.
08:33The patient can benefit from lobectomy.
08:35So that was a big change and all these
08:38were incorporated in the 2017 Bethesda.
08:41So now after the publishing
08:44publication of the 2017,
08:46we realized that still we have new
08:49data which was coming in based on
08:52newer prospectively analyzed studies.
08:54Molecular testing although was
08:56available in 2017, the menu expanded.
08:59So now we have much more complex
09:01elaborate molecular tests not only
09:04providing information and diagnostic but
09:06prognostics and therapeutics as well.
09:09And as I mentioned the ATF guidelines are.
09:11Being revised later this year,
09:14there is some category named
09:16revision in 2023,
09:18the Tessa and we all know The Who Blue
09:20book for histologic classification of
09:22tumours just came out and we wanted
09:25to incorporate those changes as well.
09:27And we wanted to make Bethesda
09:30reporting system more global,
09:31although it's global as such.
09:33But we wanted to invite more
09:35international authors and all
09:37of these led to us rewriting the
09:40Bethesda system which
09:41will be. All the 2023 that has the system.
09:44I know a lot of you will be attending
09:46the International Congress of
09:48Psychology in Baltimore next month.
09:50I would be moderating a special session
09:53called Quest for Perfection Sack title,
09:55which I'm presenting to you today.
09:57And we'll be discussing the changes that we
10:00will be implementing in the 2023 the tester.
10:03So this is where we are
10:05heading with the 2023 Bethesda.
10:07One change you will notice is that doctor
10:10at Seabus is no longer the Co editor.
10:12I think, he said.
10:14Trying to make life easier now,
10:16so I Paul Vanderlaan,
10:18who's a very well known psychologist
10:20in the Howard system, will help me.
10:22I Co edit the book and since the book is
10:25expanded a little bit with two new chapters,
10:28I needed some good associate editors.
10:31So bad blows from Upenn theatrics
10:34go shaft relay from France,
10:36Fernando Schmidt from Portugal
10:38and Philip Beal from Paris,
10:40France.
10:40All well known cytopathologist with
10:43expertise in thyroid will.
10:45Will assist US associate additives.
10:48We will have a complete endorsement
10:51by Yorgan Federation psychologist
10:53societies authorship is increasing to 65.
10:56There's becoming more complex.
10:58For the first time we are involving
11:00more clinicians and radiologists.
11:02The two new chapters would be one on
11:05molecular testing and ancillary studies and
11:07one on clinical perspectives and imaging.
11:10As I mentioned,
11:11the risk of malignancies would be revised.
11:14Pediatrics thyroid disease is managed
11:16very differently and if you look at their.
11:19ROM's,
11:19they are very different from adult patients.
11:21So for the first time we are
11:25addressing the issues of Rome's in
11:27pediatric patients and the management
11:29plan which is quite aggressive and
11:31I'll show it to you in a minute.
11:33The other thing for the first time
11:35we were able to convince everyone
11:37involved with the business system
11:39to have a single name for the
11:41six diagnostic categories.
11:42So I'm happy to report that we are keeping
11:45non diagnostic unsatisfactory is out,
11:48we are keeping a USB.
11:49Which is a much more acceptable
11:52term than plus,
11:53which is out and suspicious for
11:55follicular neoplasm is out,
11:56and it will simply be follicular neoplasm.
11:59So I think this is a major,
12:01major change the way the six
12:04terminologies will be used.
12:05I strongly believed when I started
12:08to work on this with Paul that
12:11AUS has to be simplified.
12:14Previously we had two big groups,
12:16nuclear and architectural,
12:18the micro follicular,
12:19and then we have a long,
12:21long list of miscellaneous that
12:23put rare plasmacytoid cells,
12:25hurthle cell, atypia.
12:26We had cyst, lining cell,
12:29racimo bodies, atypical lymphocytes.
12:31So studies have shown that there
12:34are two distinct.
12:36Were low risk and high risk,
12:37and I'll mention those in
12:39more detail in a minute.
12:40We also wanted our reporting system
12:43to be in alignment with The Who
12:45classification and happy to report
12:48that although the publication
12:49date would be June of 2023,
12:51we are way ahead of the scheduled
12:53time and we should definitely
12:55have the box sometime in spring of
12:57next year. So in most simple words
12:59this is how it would look like,
13:01as I mentioned non diagnostic
13:03benign AUS which is a much more
13:06flexible term accepted by a majority
13:08of people as opposed to floss.
13:11Because the the term follicular
13:13sometime would not cover the atypical
13:15lymphocytes or atypical cyst lining
13:18cells or atypical plasmacytoid cells.
13:21When you think of medullary carcinoma
13:23follicular neoplasm would stay.
13:25But we will definitely recommend
13:27to add a statement that 1/3 of
13:29the cases which are diagnosed
13:31floccular neoplasm and follow up
13:33turn out to be hyperplastic nodule.
13:35No matter how good you are and experience
13:37you are and thyroid psychology
13:39and SFM and malignant would stay.
13:42So now this is a table which
13:45essentially shows you the 2010 in
13:47Red 2017 and the in green the 2023,
13:52the upcoming risk of malignancies.
13:54And for the first time we have also
13:57calculated the average for for each group.
13:59So as you can see there is some
14:02revision and especially with with the
14:06benign ones slightly up to 6% from
14:103 and some change in the management.
14:12Profile,
14:13again a more conservative approach
14:15because we have seen that the low
14:19risk AUS the one which has which is
14:22based on micro follicular architecture
14:23or or other miscellaneous features
14:25as opposed to the nuclear one
14:27carries a lower risk.
14:29And if patient does not have suspicious
14:31imaging or clinical findings,
14:32a lot of times they don't do even,
14:34don't even repeat the efinancial
14:37animations and patients simply
14:39followed up by clinical examination
14:41or possibly with the addition of a.
14:43Repeat FNA,
14:44but diagnostic lobectomy is
14:46something which is strongly
14:48discouraged follicular neoplasm,
14:50again molecular testing is taking
14:53the the the front seat when it comes
14:56to working out these patients since
14:58we now have targeted treatments
15:00available and it's included in
15:02the manual of most of these.
15:05Commercially available molecular
15:06tests for the first time,
15:08the recommendation is to only to
15:11also do it in SFM and in some
15:14cases of malignant as well.
15:15So these are the pediatric stroms,
15:18the green one is would be for adults and
15:21the red one are for pediatric patients.
15:24And you can see there are significantly
15:27higher for example a US 50% upper end,
15:30100% for political neoplasm,
15:33near 100% for SFM and malignant
15:35and this is what happened.
15:37They are treated very aggressively
15:39with most of the time these patients
15:41are going for a total thyroidectomy
15:44or at least lobectomy.
15:46So this would be adequately
15:48covered in in the new edition.
15:50So as we all know that the World Cup of
15:52thyroid patient would would start with
15:54the looking at the clinical features,
15:57someone will palpate an audio in
16:00in a patient and then the patient
16:02is sent for ultrasound.
16:04So clinical features and imaging
16:05finding would be the pre afna workup
16:08and then based on if there is an
16:10indication to do an FDA patient will
16:12get an FNA cytopathology these days
16:14most of the time with the additional
16:17molecular analysis.
16:18So this is the most important.
16:20One part of the equation,
16:22because you will be the one deciding
16:24whether patient should be left with
16:27for clinical follow-up alone or needs a
16:29lobectomy or near total thyroidectomy.
16:31Again to emphasize that all thyroid
16:33hyphenates should be done under image
16:36guidance if imaging is available.
16:38And these days every institution
16:40has ultrasound available.
16:41As I said, seeing,
16:43seeing is believing you can see the
16:45tip of the needle when the FN is
16:47done because it appears echogenic,
16:49you know for sure.
16:51Where you are painting the sample
16:53and the other important thing would
16:55be to stay away from vessels which
16:58you can do often with palpation only.
17:01And also it's interesting fact very
17:04widely published at 50% of the time
17:06when you do image guided FA you
17:09discover more significant finding
17:11other one or several new nodules
17:14or you discover nodes in the neck
17:17which assume a lot of importance in
17:19a patient with a with a primary.
17:21Other nodule.
17:22So it's important that imaging
17:23should be done in all cases which
17:26are going for thyroid affinity
17:28and ideally the thyroid affinity
17:30itself should be done with imaging.
17:32The other fact is that one of six
17:35patients when you do image guided
17:37fnas you will not find the nodule.
17:40And I often give my own example.
17:42A couple of years ago my internist
17:44discovered a nodule in my thyroid and
17:46and he knows that thyroid is my main area.
17:49So smilingly said off you go.
17:51Parathyroid F&amp;H.
17:52So I showed up in ultrasound as a
17:54patient and the radiologist who was
17:56doing the FDA could not find the nodule.
17:59So that really takes out a lot of burden,
18:03a lot of pressure from the patient
18:06and clinician because then there
18:07is no need to do an FHA if there
18:09is no nodule present.
18:10So these are important factors
18:12which should convince everyone
18:14that all thyroid affinity should
18:16be done under ultrasound guidance
18:18and potentially you can also locate
18:20complications which are.
18:21Yeah,
18:22like hematoma formation here you can
18:24see between the calipers a isoechoic
18:26hematoma you can actually see if
18:28you got and the colour Doppler
18:30on the source of the bleeding.
18:32And these patients need
18:34conservative management,
18:35just cold compresses that
18:37will treat the hematoma.
18:39So non diagnostic.
18:40As you can see it's not a trivial
18:43diagnosis because it does carry a
18:46significant risk of malignancies and hence
18:49these cases should always be repeated.
18:52And ideally like I said
18:54under ultrasound guidance.
18:55The contentious thing around 2017
18:57was which we actually corrected in
19:002017 edition was what should be the
19:02time interval between the first non
19:04diagnostic F and A and the repeat
19:07FA and everyone agreed in 2010.
19:09There should be 3 months or 12
19:11weeks because you want to give
19:13thyroid a chance to heal.
19:14The trauma of the previous FN,
19:16which in 2017,
19:17based on several series which
19:19were published was proven wrong.
19:21You can actually bring back
19:23the patient much sooner.
19:24At my institution I often get calls
19:26from radiologists when to bring the
19:28patient back and I always tell them
19:30three weeks or four weeks would be ideal.
19:33A patients would show up for repeat FA.
19:36Three months is a long time
19:38to put for the patient.
19:40To undergo the agony of waiting.
19:42Sometimes they don't even show up
19:44after such a long period of time.
19:46So again the conclusion in 2023
19:48Bethesda which you will see is that
19:50you can have much shorter interval
19:53to repeat the FDA in terms of the
19:55incidence of non diagnostic 10%.
19:57Still the cut off that not more than
20:0010% of all your thyroid FDA should be
20:02non diagnostic was higher than 10%.
20:04You have to modify the technique which
20:07typically based on too much blood.
20:10And very few cells.
20:12So modify the technique,
20:13shorten the needle dwell time
20:15to 6 seconds of faster,
20:17use a thin needle and Hopkins we use
20:19a 26 gauge ideal needle and also we
20:22don't apply suction which is common
20:24sense thing because thyroid is so vascular.
20:27Adequacy criteria remains the same
20:29although we had a very strong urge
20:32to change it from six group with 10
20:35cells each or roughly 60 cells and
20:38almost till the chapter was finalized.
20:41On adequacy,
20:42we wanted to revise it to say three
20:44groups of four groups were unfortunately,
20:46although most people agree that
20:48it should be revised to a much
20:51smaller count of cells,
20:52there are no validation studies available.
20:55So then in the end we concluded that the
20:58criteria will remain the same as 2017,
21:00although if it's clinically benign appearing,
21:04imaging benign appearing,
21:05you could certainly call something benign
21:08with fewer than six groups of cells.
21:10The contentious issue of cyst food only
21:14remains for 2020 with test as well.
21:17For example,
21:17here you see a 10 Volt cyst which is
21:20almost an aquatic because the ultrasound
21:22waves are passed through and through,
21:24they assist through it.
21:25Not no protein,
21:26no solid component.
21:27What to call these cases and when
21:31you look at it under the microscope
21:33you will see these hemosiderin laden
21:36macrophages with these big clear zones on
21:38the slide signifying 10 watery colloid.
21:40Which will not stick to the to the
21:42slide but no follicular epithelium.
21:44So we again emphasize that you
21:47call these cases non diagnostic,
21:51but you will write a note to the
21:53clinician that if there are no
21:55suspicious clinical ultrasound,
21:56finding them will nodule no
21:57calcification in the world,
21:59no irregularly thickened wall,
22:00so on and so forth.
22:01These cases could be deemed
22:03adequate and benign,
22:04but the decision should be with the
22:06clinicians from your end you will
22:08still call them non diagnostic system.
22:10Wrongly, and the reason is simple.
22:13For example this this nodule was
22:15done under palpation guidance and
22:17hence the cystic part was of an aid.
22:19The mural nodule and the solid
22:22component was not FN 8.
22:23Typically these are complexes
22:25with solid cystic areas,
22:26but look at these punctate echogenic foci.
22:30Translating, translating to my
22:32calcification or aggregates of some
22:34bodies are very specific imaging
22:36sign of apparitia carcinoma and
22:39when you aspirate such cases.
22:41Again you will end up with
22:42a lot of cyst fluid,
22:44very transparent zones on the slide,
22:46hemosiderin laden macrophages.
22:47And if you look at these cells,
22:50which look very histiocytic because
22:52of their cytoplasmic appearance,
22:54the nuclei are not those of histiocyte,
22:57these are over,
22:58they show grooves and in some
22:59cases you will see these beautiful
23:02intranuclear inclusion despite the
23:04fact that they look mysterious sick.
23:06So this is a book picture of cystic
23:09variant of papillary thyroid carcinoma.
23:12Sometimes people who do only
23:14histopathology as the question,
23:15well, WHO does not endorse the
23:18cystic variant of papillary carcinoma
23:20where we as syrup pathologists and
23:22radiologists like to consider it
23:24separately because it helps us with
23:27our diagnosis and differential diagnosis.
23:29We try to exclude the possibility
23:31of cystic change in an abnormal
23:33roid nodule when we talk about
23:35cystic variant of papatya carcinoma.
23:37And just to let you know,
23:38I have never seen intranuclear
23:40inclusions in histiocytes.
23:42So the moment I start to see
23:44Intranuclear inclusion and histiocytic
23:46cells in a thyroid FNA,
23:48I think of parathyroid carcinoma
23:49and these are further examples.
23:51Even on air dried beef which stain smear,
23:54you will see beautiful intranuclear
23:56inclusions and these cells imbibe fluid
23:59when they are in fluid environment
24:00for a long period of time and they
24:03begin to look like so bubble cells
24:06the classic appearance but because of
24:10the nuclear features it's it's papillary.
24:12Liquid based cytology further
24:14complicates the issue because everything
24:16looks ascitic in these cases.
24:18So that's a histiocyte.
24:19Histiocyte where most of these cells
24:21with beautiful internuclear inclusions
24:23are cystic variant of Appalachia carcinoma.
24:27They're simply undergoing this
24:28phenomenon what is called hyper
24:31regularization that they absorb fluid,
24:33develop vacuoles and begin to
24:35look like histiocytes.
24:36Benign,
24:37very satisfying category because
24:38we know the risk of being agency
24:41is low these four entities.
24:43Are diagnosed on FNA.
24:44The most common one we know
24:46is adenomatoid nodule.
24:47Again the same recommendation
24:49that these patients do not need a
24:52repeat FNA or a biopsy or anything,
24:54simply followed up by clinical
24:56with or without imaging follow up.
24:59We all know that imaging helps.
25:01This is a number.
25:03Nodule is a disfiguring disease.
25:04Multinodular, often bilateral,
25:06often can have eggshell calcification
25:08and gross examination.
25:10We have all seen how it looks like.
25:13It's very cellular on psychology
25:15because it's a hyperplastic process.
25:17I hate to add the word cellular before
25:20adenomatoid nodule because it's meaningless.
25:22They're all cellular.
25:23Sometimes they are.
25:25They are macro.
25:26Follicles are gigantic follicles,
25:27and when you repeatedly.
25:30Puncture them with an FNA,
25:31you will disrupt the wall and they
25:33will become flat monolayers but with
25:36abundant colloid in the background.
25:37So reassuring finding if you
25:41see abundant colloid.
25:43The other thing which you would
25:44sometimes see would be over
25:46abundance of colloid and again
25:48your radiologist will help
25:49you. For example in this case it's a
25:52multi septated nodule looks anechoic
25:55but it has these four side of an
25:58artifact called Comet Tail artifact.
26:00They have small tails and this typically
26:03happens when the ultrasound waves hit
26:06a protein rich fluid like colloid.
26:08It creates this artifact
26:10reassuring finding and if you see.
26:13Even a lot of cells, such as in this case,
26:16you know that you are dealing
26:18with a benign diagnosis.
26:19So it's an easy diagnosis,
26:21very specific diagnosis.
26:22Very few cases will have malignant tumor,
26:26but there are issues.
26:27We all know if overabundance of composites
26:30can lead to an oncocytic neoplasm diagnosis.
26:34If your tip of the needle hits
26:36an area with all micro follicles,
26:38you can over diagnose it
26:40as follicular neoplasm.
26:42Sometime with cystic component
26:43you can have cyst lining cells
26:46that look fairly atypical.
26:47Occasionally with cytoplasmic
26:48tales I I shared this case with
26:51the resident this morning.
26:52This famous metaplasia,
26:54you can get intranuclear
26:56inclusions and of course capillary
26:58architecture can confuse you with
27:01the appellate higher carcinoma.
27:03So this is my favorite slide to
27:05show and remind people that the most
27:08common cause of a false positive
27:10diagnosis in thyroid cytology is
27:12Hashimoto thyroiditis over diagnosed
27:14as suspicious for papillary papillary.
27:16So if you're not careful,
27:18you can have a bad diagnosis as
27:20far as long as you remember that
27:23you have to be very careful with
27:25Hashimoto again imaging can help.
27:27It's a bilateral symmetrical process,
27:30it's a lobulated.
27:33Lobulation that you will see on imaging
27:36micro mobilizations because of these Gray,
27:39white lymphoid nodules and very
27:41often the estimate would be widely
27:43dilated as you can see here and in
27:46the initial phases of Hashimoto.
27:48If you do color Doppler it's
27:50very vast or it's late stages
27:51that you lose the vascularity.
27:53Sometime adding up these features
27:55even before you get DF and A
27:57will will give you a good idea
27:59that is going to be a lymphocytic
28:01thyroiditis and if you are lucky you.
28:03We get this beautiful biphasic
28:05admixture of hurthle cells
28:07and polymorphous lymphocytes,
28:08which may appear singly or in the form
28:11of this lymphohistiocytic aggregate,
28:14as I'm showing you here.
28:16Problems usually arise when you
28:18have reactive changes in the
28:20nuclei of the metaplastic cells.
28:22So for example here the background
28:25is lymphocytic.
28:26These cells have markedly enlarged
28:28nuclei which are over in shape.
28:30Look at this case.
28:32All nuclei with beautiful grooves,
28:34intranuclear inclusions,
28:36fine powdery chromatin Oval nuclei,
28:39beautiful intranuclear inclusions.
28:41So it's very hard to suppress the
28:44temptation not to call these three
28:47scenarios or cases as suspicious of PTC.
28:50PTC for all three cases on resection
28:53was simply Hashimoto Paraditas.
28:55So what I'm trying to tell you is
28:57that have a very high threshold
28:59for committing to PTC.
29:00There's a special BDC diagnosis when
29:03you have previous history of Hashimoto
29:05or use seem concurrent changes of Hashimoto,
29:08even with the knowledge that patients
29:10who have long standing Hashimoto
29:11have a much higher incidence
29:13of typhoid higher carcinoma.
29:15Just tell yourself that you're
29:16not going to call it PTC unless
29:19you're absolutely convinced,
29:20so I'm not stopping you from doing that.
29:22In my case,
29:24when I see such focal changes,
29:26I call it a US nuclear tipi in a
29:29background of Hashimoto.
29:31Or I go up to suspicious for PTC if several
29:34slides show internuclear inclusions.
29:37But if I have edge to edge carcinoma,
29:40those are the cases where I feel
29:43comfortable calling TTC rising in the
29:45background of Hashimoto's thyroiditis.
29:47So you all know AUS has a lot
29:50of subjectivity involved.
29:51There are many phases of this diagnosis.
29:53So these are the cases which are
29:55typically have very subtle focal atypia
29:57that you don't want to call benign
29:59and lose the patient to follow up.
30:01On the other hand,
30:02you don't want to go all the way to
30:05Flagler Neoplasma SFM and have the
30:07patient undergo polypectomy or thyroidectomy.
30:09So you use the term AUS.
30:12And now this is the major change
30:13that you will see now are based
30:15on our own study at Hopkins.
30:17We showed that there are two
30:19distinct groups of AUS.
30:20The one which are based on nuclear
30:23tipiya has an ROM of 48% in our study,
30:2623% of all other subtypes of AUS.
30:29And now we have support of several
30:32molecular studies which have clearly
30:34shown that the high risk group have
30:37mutations more specific like B RAF
30:39V600E and the other group which we call
30:41micro follicular.
30:42Or miscellaneous have the low risk
30:45mutations such as KSP 10 taxi PPR
30:48gamma pointing more towards nifty or
30:51flocculant neoplasm type diagnosis.
30:54So that prompted us to have in
30:572023 editions simply two AUS,
31:00AUS nuclear this is the highest and
31:03AUS other which is the low risk.
31:05If you want you can certainly
31:07put a descriptor after your
31:09diagnosis based on micro,
31:10follicular or plasmacytoid cells or cyst.
31:13Lining cells, but the top line would be AUS,
31:16nuclear or AUS.
31:17All other benchmark is the same
31:20and not more than 10% of your
31:23cases should be called the US.
31:26If it's significantly higher,
31:27you have to bring it down.
31:30Polyclar neoplasm,
31:30again imaging may help.
31:32They typically oven in shape.
31:33They could be ISO,
31:35equate or high bit hypoechoic.
31:38They're very vascular when
31:40you initially read them,
31:41typically very irregular
31:43vascularity in the middle.
31:45This is a a case which turned
31:46out to be Flagler Red Norma.
31:48You can actually see the needle track
31:50and hemorrhage caused by the FNA.
31:51This turned out to be a minimally
31:53invasive follicular carcinoma.
31:54You can see the disruption
31:56of the capsule here,
31:57but these features cannot be seen.
32:00Accurately on imaging and
32:02certainly cannot be seen on
32:04psychology and even in Histology.
32:06You really have to look hard for
32:08this penetration of the capsule with
32:11mushrooming of the tumor into the
32:13lymphovascular spaces before you
32:15will commit to the diagnosis flula
32:17carcinoma and hence we simply call
32:20them cellular neoplasm and then
32:22they do a diagnostic lobectomy.
32:24These days molecular tests are done
32:261st and if they are negative within 96%.
32:29Very predictive value.
32:30You avoid the need to do
32:33a diagnostic lobectomy.
32:34A very important thing happened in
32:372017 edition which will remain in 2023,
32:40which is how to diagnose or screen
32:43for nift P and the answer is simple.
32:46The cases where you see mild nuclear
32:49changes with the philar based
32:52architecture call them follicular neoplasm.
32:55In 2010 we recommended those cases
32:57should be bumped up to to suspicious of.
33:00Primary carcinoma,
33:01but in 2017 and in 2023 Bethesda
33:04we allow minor nuclear changes
33:06to be present in follicular
33:09neoplasm because we want to put
33:11all potential cases of nift P in
33:14follicular neoplasm categories
33:15so they get the right treatment
33:17which is lobectomy for next week.
33:19So it's very important to realize.
33:20So if I would see a curricular pattern
33:24lesion with rare intranuclear inclusions
33:27or groups or some over nuclei.
33:30I would call it follicular neoplasm,
33:32even knowing that potentially this
33:34case will turn out to be nifty.
33:36But I'll write a simple note.
33:37Differential diagnosis includes an FPN
33:39collision that happy surgeons are happy for.
33:42Lobectomy will only show a nifty.
33:46So that's the psychology
33:48of cellular neoplasm.
33:49Again the criteria remains the same
33:51that it should show a predominantly
33:54micro floccular architecture as
33:56shown here that more than 50% of
33:58the cells will be in the form of
34:02these very uniform equal size micro
34:05follicles with 15 or fewer nuclei in
34:07the form of a complete circle or near
34:11complete circle with no hole oil in the
34:14middle of the follicles they can have.
34:16Get your bachelor architecture.
34:18But whatever the case can colloid
34:20more than 50% of popular component
34:22in micro follicular architecture.
34:24You will call it follicular neoplasm.
34:27Some change changes happen
34:29with Hercule cell neoplasm.
34:31Again, this is a diagnosis that
34:33you will base on near exclusive
34:36population of further cells.
34:38I know in Histology it could be 70%
34:41or whatever percentage cut off.
34:43People would decide,
34:44but it's inside a path we like
34:46to see near 100% of the cells
34:49as appearing as ankle slides.
34:51The other change which happened
34:52with this category is that as you
34:54know WHO junk the term hurthle
34:56cell and everyone is happy.
34:58That they're simply call
34:59on cosmetic neoplasm.
35:01So using Bethesda system,
35:03perspectively speaking,
35:04you will simply call them
35:06oncocytic follicular neoplasm.
35:08A very simple diagnosis are very
35:10commonly asked question is how
35:12to differentiate hurt a oncocytic
35:15hyperplasia as part of a hyperplastic
35:18nodule from a true oncocytic neoplasm.
35:212 features that I pay.
35:22Due importance would be the proliferation
35:25of capillaries as shown here.
35:28And frequent bind nucleation.
35:29If every other cell has two nuclei,
35:31especially with the nucleolar prominence,
35:34I would be very comfortable calling
35:38it oncocytic follicular neoplasm.
35:40A major differential diagnosis
35:42of a floccular neoplasm.
35:44Is with a parathyroid lesion typically
35:48a parathyroid Noma rule of thumb,
35:51which I usually teach my fellows,
35:54is that if you're going after
35:55a hard F and a nodule and you
35:58have zero call or no call roid,
36:00think parathyroid medullary
36:02and metastasis in that order.
36:04And certainly intra parathyroids
36:06could be intra thyroidal.
36:08Even the best radiologists cannot
36:10distinguish them and we will come to
36:12your attention as as a thyroid nodule.
36:14So they are very cellular because
36:16they are stronger poor tumors.
36:18They have a diffuse population of cells,
36:20some of which may form follicles,
36:22but large population of naked
36:24nuclei because when you smear
36:27them you lose the cytoplasm.
36:29Very vascular as you can see here.
36:31So at Hopkins what we do is that
36:33we keep this small bullet tubes
36:36plastic tube prefilled with one CC
36:38of Hanks balanced salt solution on
36:40the FSA card and whenever we suspect
36:42something could be a parathyroid.
36:44That's the radiologists do one extra
36:47dedicated task and then we rinse
36:49that pass in this one CC of hand
36:51down substitutions, Andrew Chemistry.
36:53And next day when the PTH level
36:56comes back higher than the patient
36:59peripheral blood that's highly
37:01protective of a parathyroid lesion.
37:03I use the term parathyroid lesion.
37:05I don't call it parathyroid neoplasm
37:07because you could certainly have
37:10isolated hyperplasia recently or you
37:12can make a cell block and do for example.
37:15PTH immunostaining,
37:16these eight people love to do GATA 3,
37:19which is an excellent marker
37:20because it's saying the nucleus.
37:22So even if you don't have much
37:24cytoplasm in a limited sample,
37:25you can still get nuclear staining.
37:27So one thing I wanted to tell you
37:30all is that recently people have
37:33started to ask this question.
37:34Well, you're calling it parathyroid
37:37lesion patient also has a neck node.
37:40Could this be a parathyroid carcinoma
37:42or is it a parathyroid adenoma?
37:44Unfortunately we have this
37:47excellent immunostain now,
37:48a pair of fibrominn which is
37:51extracted from a tumor suppressor gene
37:55hyperparathyroidism 2 HPT two which?
37:58Which is lost in parathyroid carcinomas.
38:00And look at the impressive profile
38:04for 96% sensitivity and 99%
38:07specificity for parathyroid carcinoma.
38:10So this is a neck node that was sent to me.
38:13There's a smear and the cell block.
38:14So we knew right away it was
38:16a parathyroid tissue.
38:18Patient also had a thyroid nodule
38:20and you were asked the question
38:22is it a metastatic parathyroid
38:24carcinoma to a net node.
38:25So we did a part of Fibrominn as you can see.
38:28Beautifully, strongly positive.
38:30So we excluded carcinoma based
38:32on this very impressive profile.
38:35They went after the thyroid nodule.
38:37It was simply a benign hyperplastic nodule.
38:40So this is a very important
38:42part of my presentation.
38:43The so-called indeterminate thyroid
38:45nodules as you know is a very
38:47frustrating diagnosis for pathologists
38:49and also for the clinicians.
38:5215 to 30% of them are indeterminate.
38:55Majority unfortunately are still resected
38:57and majority do turn out to be benign.
39:00So it's an unnecessary procedure.
39:03AUS and political and neoplasm are
39:05typically taken as as a categories
39:08within the indeterminate diagnosis.
39:10I love to show this slide because
39:12now there are molecular tests which
39:15people are afraid will replace
39:17cytopathologist or thyroid FN's.
39:19Just like HPV stole a lot of GYN pathology.
39:22So I often show this slide that you
39:24will not get to this stage that you
39:27will find work for food because you
39:29will always have your thyroid cytology.
39:32These tests are expensive so they
39:34will not be offered to every patient.
39:36With a thyroid nodule you will
39:39still do your critical role.
39:41You will do the FDA,
39:42read the FN as in determine it and then
39:45the patient will get a molecular test.
39:47But we all agree that they do
39:51increase significantly preoperative
39:53diagnostic accuracy and FN sample,
39:55the liquid sample is a great sample
39:58to do these molecular testing.
40:00So equation will always flow this
40:02way that we will have an affinity
40:04diagnosis leading to a molecular test.
40:06And again these molecular tests as
40:09I mentioned before reduced the.
40:11Number of so-called diagnostic
40:12lobectomies which we used to do,
40:14again not enough time to go into
40:16details of these tests.
40:17We all know that there are three major
40:19tests that are often used in pharma, tyroc,
40:22Dynex, all three are very comfortable.
40:25At Hopkins, we use a pharma.
40:28Title seek and Thai jennex provide
40:31an additional advantage that you
40:33can do on paraffin block or you can
40:36escape the cells off or glass slide.
40:38But keep in mind they do offer a
40:41much lower success rate when you
40:44don't do it on on a liquid sample.
40:47So whichever test you do they all give you
40:50good results and this is a very important.
40:54Editorial which was written by.
40:57The well known endocrinologist who had helped
41:00us tremendously with the Testa book also.
41:03And as you can see here that he's clearly
41:06telling the readers that these tests will
41:09not replace what we do the imaging and FNA,
41:12it would simply support the FN a
41:17diagnosis or a clinical diagnosis
41:19or higher diagnostic accuracy.
41:22So we are looking for a synergistic
41:25partnership with these molecular tests.
41:27They're not here.
41:28To replace us or replace the radiologist.
41:32So suspicious malignancy,
41:33no major change.
41:35But just to remind you that the cases
41:38previously what we used to call
41:40suspicious for PTC based on focal
41:43nuclear features with fine powdery
41:46chromatin or grooves or rare intranuclear
41:49inclusions now should be called AU for,
41:52should be called follicular neoplasm
41:54because these are potential cases
41:55that will turn out to be nifty and
41:58you want these patients to get a
42:00lobectomy and not a total thyroidectomy.
42:02So that's the only message I have for
42:05you when you entertain that diagnosis,
42:07suspicious for malignancy,
42:08that be conservative for lesions
42:11which have a follicular pattern and
42:13all these showing focal features
42:16of papillary carcinoma.
42:17So malignant remains a very accurate
42:21diagnosis in Afghani psychology,
42:24no significant change from 2017.
42:28If you are not seeing the major
42:30diagnostic features of papillary
42:32fibroelastoma course small bodies
42:34which are seen in minority of cases,
42:37in my experience not more than 1520% of PTC.
42:40So those features,
42:42ask yourself question should I be calling
42:44it PTC or should I be happier calling?
42:48Suspicious of PDC if it has a follicular
42:52pattern architecture if you are calling it.
42:58PTC and it has a follicular pattern or
43:01as a matter of fact in all PTC cases.
43:03Make it a habit to add this optional
43:06note at three to 4% of the cases which
43:10are called PTC on affinity psychology.
43:12Turn out to be nifty or follow
43:14up to protect yourself,
43:15protect declination and so that you
43:18will still have a good confidence from
43:21your clinical colleagues in case it
43:23turns out to be nifty and patient,
43:26get overtreated.
43:28We all know how papillary carcinoma
43:30looks like.
43:30The workout starts with imaging isoechoic,
43:34beautiful punctate,
43:35echogenic foci,
43:37cluster of microcalcifications,
43:39similar bodies,
43:40beautiful papillary architecture,
43:41fibrovascular course.
43:42And now I can see that I did not
43:45see show the fellows the classical
43:47fibrovascular scores earlier this morning.
43:50But here you go. This is picture perfect.
43:53All nuclei with longitudinally
43:54running groups,
43:55tiny marginal nucleoli sitting
43:57underneath the nuclear membrane.
44:00And of course sharply punched
44:02out intranuclear inclusions,
44:03classic features of papillary.
44:05So if you see this,
44:07it's simply a papillary thyroid
44:10carcinoma diagnosis question.
44:11Often asked is should we be diagnosing
44:14variants of subtype and the answer is no,
44:17because sometimes even if you have these
44:19tall cell variants which you are seeing here,
44:21the tallest cells,
44:23the oncocytic look so bubble
44:25internuclear inclusions
44:26or you're seeing these gland
44:27like formations reminding you
44:29that this could be a colonic.
44:31No carcinoma. These are columnar
44:34salvadorians or capillary.
44:35I think these changes are very
44:37focal and you may be inaccurate
44:40when there is action is done.
44:42So parathyroid carcinoma
44:43is a great diagnosis.
44:45I only mention it in a note
44:47that's focal features suggest that
44:50all sylvarant or the very rare
44:52columnar cell variants if I see
44:55those features because sometimes
44:57it will help the surgeon do the
44:59central level 6 of paratracheal.
45:01And for the section,
45:03because that nodal dissection is
45:05associated with the two serious
45:07potential complications I have surgery
45:09the permanent laryngeal nerve damage
45:12and or permanent hypoparathyroidism.
45:14Most surgeons will take those
45:16node out nodes out anyway,
45:17but good surgeons palpate them.
45:19If it's negative image
45:21them imaging is negative,
45:23then they may depend on molecular
45:25testing or your call of tall cell
45:27variant or an aggressive variant.
45:29But even with that I strongly
45:30recommend that don't suck.
45:32Like that with her carcinoma.
45:35Medullary is an easy diagnosis,
45:36but also keep in mind it has the widest
45:39spectrum of cyto morphologic changes.
45:41Very often it will be the so-called
45:44epithelioid type with plasmacytoid
45:46features could be focally spindly.
45:49The presence of amyloid greatly helps,
45:52but you have to learn how to recognize
45:54amyloid as opposed to Polaroid.
45:56It's more glassy appearing,
45:58more amorphous, more sharply defined,
46:01cut out fragments, and if you recognize it.
46:04Which is present.
46:06I think in 6570% of the cases
46:09it becomes an easy diagnosis.
46:11But issues happen when you have a very
46:14uncommon pure spindle cell morphology
46:17and you cannot exclude spindle cell lesion.
46:21Or someone asked me this morning
46:22at the 8:00 o'clock lecture,
46:24how about the giant cell variant
46:27of megalithic carcinoma.
46:28Very polymorphic cells still
46:31retaining a dark,
46:33hyperchromatic,
46:33evenly dispersed chromatin.
46:34So if you are on site for efna in
46:37these cases now we recommend that
46:39you can actually send the needle.
46:41Watch out for calcitonin levels.
46:44You don't even have to wait for a cell
46:46block and if the calcitonin comes
46:48back significantly higher than you know,
46:50you have a measure tire carcinoma.
46:53Previously we recommended it on
46:55potential metastasis to the lymph
46:58nodes of a majorly carcinoma Ki 67
47:00with now WHO recommends be done
47:03for to classify low and high grade.
47:05But it's a histologic thing and I strongly,
47:08strongly recommend do not attempt
47:10it on psychology.
47:11Just the simple diagnosis of measuring
47:14high carcinoma should be good enough.
47:16And a plastic is a fairly easy diagnosis
47:19when you see it in psychology.
47:22As the name suggests,
47:24extreme pleomorphism or an aplasia.
47:26You can imagine that disparity
47:28in cell size and shape.
47:30So a very easy diagnosis.
47:32And we all know it's a rapidly enlarging
47:34mass often cause airway compression.
47:37This patient died inhaling
47:39the necrotic tumor.
47:41Imaging is a very aggressive looking
47:44tumor with necrosis extra thyroidal.
47:47Sanction a significantly enlarged
47:50and juxtapose cervical nodes.
47:53But on psychology,
47:55sometimes it's very difficult
47:57to exclude metastasis.
47:59I often tell people when you have a
48:01pleomorphic tumor and you are trying to
48:03jump quickly to a diagnosis of anaplastic,
48:05always exclude metastasis
48:07and some oncologic centers.
48:10Metastatic carcinomas are cancers are
48:12more common than primary anaplastic,
48:15for example.
48:15This case looks very similar to
48:17the one which I'm showing you here,
48:19except this is a metastatic Melanoma in a
48:22patient who had a known history of Melanoma.
48:24So if you don't get the history or you
48:26don't pay particular attention to history,
48:28you can very easily misdiagnosed metastasis.
48:31So history is important and certainly keep in
48:35mind that renal cell metastasis lung adeno,
48:38breast add no Melanoma,
48:40these are metastasis to thyroid
48:42traders are very vascular organs
48:44and it knows very effectively how
48:46to filter circulating tumor cells.
48:48So it's not uncommon to find
48:50metastasis in the thyroid.
48:52We all know when you've
48:54anaplastic thyroid carcinoma.
48:54You will lose sight of carrot in staining.
48:56Often you will lose sight
48:59of globulin and TF11 stain,
49:01which is often retained at least focally.
49:03Is this beautiful pack stain immunostaining?
49:06So still it remains the most
49:09important immunostains to
49:11diagnose anaplastic carcinoma.
49:13Also to remind you that the new WHO
49:16classification took out squamous
49:18cell carcinoma as a separate entity.
49:20Now if you see squamous cell
49:22carcinoma in a thyroid,
49:23either it would be a locally
49:26invasive or amacs famous cell or
49:28it will be a subtype of anaplastic.
49:30So if I see this prospectively
49:32in in a thyroid FNA,
49:34I would call it anaplastic undifferentiated
49:36paracast Noma squamous cell type.
49:38That's how we will report it.
49:41I'm not going to mention too much about nift.
49:45Pi will simply answer 2 questions for you.
49:49Is there a cytopathologic
49:50diagnostic criteria for Nipp?
49:52And answer is simple none.
49:54Should we be diagnosing if PN,
49:56FN,
49:57cytology and answers absolutely no?
50:00It's just that we seen for nifty and
50:04in cases that we think are potentially
50:07nifty based on focal papillary features
50:10and follicular pattern architecture,
50:13we will put them in follicular neoplasm.
50:15If we like we can add a node
50:19differential diagnosis including FP.
50:20Most importantly,
50:21the understanding is that it leads
50:24to de escalation of treatment,
50:27reduces the number of diagnostic lupa.
50:30Surgeries, especially total thyroidectomy.
50:33So these patients prospectively when
50:35you call them should only undergo a
50:38lebec me and not a total sorry text me.
50:42So we all know the critical role of
50:45molecular testing and diagnostics,
50:47but now it has expanded into.
50:51Agnostics and precision or targeted
50:53medicine as well.
50:54So not only emerging but established
50:58evidence between genomic variants
51:00and tumor Histology how they behave,
51:03what would be the therapeutic clinical
51:05course and therapeutic options available.
51:08So really A3 prong role for molecular tests.
51:12For example,
51:13most of the commercially available
51:15molecular tests now do dirt,
51:16ALK and entrec analysis and all
51:19these mutations are associated with.
51:21All the bad things you can see in a
51:24thyroid cancer, higher mortality,
51:26clinical stage treatment failure,
51:28metastasis so on and so forth.
51:30So it does give our clinicians very useful
51:33piece of information for example entrec
51:36mutated rearranged reputar carcinoma.
51:39We all know these patients present
51:41with the with multi focal disease,
51:45aggressive disease,
51:46advanced clinical stage and unfortunately
51:49on FN sometimes you don't see.
51:52Very well developed features of papillary
51:54carcinoma and has molecular tests are
51:57very important that if they demonstrate
52:00and track rearranged carcinoma.
52:02So this is what has happened that FDA
52:05has approved a number of agents for
52:08treatment of these mutated cancers.
52:10For example,
52:12the debriefing with the additional
52:15traumatic for B RAF mutated and apothecary
52:19carcinoma larotrectinib is now very
52:22commonly used for Amtrak fused tumors
52:25regardless of the tumor Histology.
52:28And the the most recent 1,
52:30the pulsating herb is used for red mutated.
52:33Measure with high carcinoma so the
52:35molecular tests play a pivotal role
52:37and and if you don't have these
52:39molecular tests at the institution,
52:41you're actually denying the patient
52:43a chance for a disease free survival
52:47based on these very specific targeted
52:50therapies which are now commonly
52:52available in many major medical centers.
52:55So very briefly,
52:56since people do ask me the role of AI
52:59or deep learning or computer aided
53:01learning in psychology, I often.
53:03Tell them.
53:04So far it has not worked because it's
53:07very challenging to apply to psychology.
53:10Psychology has color as opposed to imaging,
53:12which is black and white or Gray.
53:15And not only color, it has many stains.
53:18Even the same stain diff quick use
53:20at Yale will look different from
53:22the TIF quake user options.
53:23The nuclei will I would different
53:26intensity of nuclear staining and
53:29computer cannot be taught that those
53:32variations because it would look for the
53:35intensity of gradient of the of the staining.
53:38Cytology has too many smears.
53:40I mean you can imagine a conventionally
53:43prepared thyroid ethnic and
53:45have anywhere from 10 to 20.
53:47Plus smears.
53:47Scanning them,
53:48especially if they're three-dimensional,
53:50adding Z, stacking,
53:51cutting them in a thin slices,
53:54using the whole slide scanner,
53:56too much data for the computer to analyze,
54:00and most importantly,
54:01I pose these questions to people
54:04who are doing AI.
54:06I'm not saying anything terribly
54:07bad about a all I'm saying is
54:10that it will take time for AI to
54:12be applicable to diagnostics.
54:14So the questions I often ask is,
54:16are you trying to replace?
54:17Colleges or societal technologists
54:20for onsite evaluation of articles.
54:23Time consuming process and
54:25most people say no.
54:27Are you trying to achieve
54:28higher diagnostic accuracy?
54:29I already presented such good risk of
54:33malignancies and accurate accuracy,
54:36high accuracy of thyroid athanas.
54:38So the answer is no.
54:39So are we trying to decrease the
54:41turnaround time and again the answer is no.
54:43Actually we will add on turn around
54:45time and we trying to save valuable
54:48healthcare dollars and answer is again no.
54:50If the companies make AI programs,
54:53they will want their money back
54:55and it will add to the cost.
54:57Of doing an FDA. So with that I
55:00and my presentation I would be very
55:03happy to answer any question about
55:05Tyler Cytopathology in general or
55:07Bethesda system in particular.
55:09So again thank you Angelique for hosting me.
55:12Thank you all for being wonderful audience.
55:21Yes, please. Problem. Quite often.
55:28Technically adequate.
55:29But they don't have enough colloid.
55:34Line. Question. But so you're
55:39struggling, you don't have microphone.
55:44Stop drinking.
55:48Or you.
55:51It's a good question and
55:52answer is very simple.
55:53If you have enough cells,
55:55that sample is not non diagnostic,
55:57it's diagnostics and it becomes
56:00an issue of interpretation.
56:02I sign out thyroid nodules benign all
56:04the time even without seeing any colloid.
56:07If the cellularity,
56:08even with the high cellularity,
56:10if the cells look benign,
56:11it's a benign diagnosis if there
56:13is lymphoid proliferation in
56:15the background even without.
56:16And if little cells I call it
56:19benign for hachimura you don't have
56:21to have sex with flicker cells,
56:23one or two groups or even without
56:25it seeing any flicker cell as long
56:27as this logical evidence or you're
56:29seeing good lymphoid population.
56:31So again answer is simple that if
56:33you have follicular cells present
56:35in a sample it's not non diagnostic
56:38unless there is a compromising
56:41factors such as air drying artifact,
56:43excessively obscuring blood,
56:44so on and so forth.
56:50Yeah. Thank you for that.
56:51Excellent. One area that.
56:57I think. It's when.
57:04Yeah.
57:09Maybe the song? Michael Pages for people.
57:16You know, but then you know,
57:16sometimes you just have exclusively
57:20about yourselves and you can see.
57:23Especially appropriate for that particular.
57:27So good question.
57:28Thank you for asking that.
57:29It always remains.
57:33An issue where to draw the line between
57:36oncocytic hyperplasia a US oncocytic
57:40type or oncocytic neoplasm and so,
57:44so answer would be that if
57:46you have adequate number of.
57:48Are on cassettes.
57:51It is not a US unless there
57:54are other factors involved.
57:56So if you have oncocytic only with
57:58lot of color in the background
58:00even without seeing any usual
58:01type of follicular cells,
58:03I will lean heavily towards a benign
58:06diagnosis because I would know
58:08that Oncocyte don't make colored,
58:10they're defective metaplastic
58:12forms of particular cells of
58:14cellular colloid is coming from
58:16somewhere where we don't not seeing
58:17the usual form of liquor self.
58:19So colloid and Oncocyte is.
58:22Is a very reassuring mixture and
58:24I would go behind the the cases
58:27where I call something a US based
58:31oncocytic subtype is all Oncocyte
58:33in a very sparsely cellular sample.
58:36Or if I see a lot of wrong quest sites.
58:41But the,
58:42the,
58:43the clinical impression is Hashimoto
58:45thyroiditis and I don't want to go
58:48all the way to oncocytic neoplasm,
58:50then I would go AUS on cosec.
58:56But again, it's it's a little bit
58:58of subjectivity is involved here,
59:00and the one advice I could give you
59:02is to do molecular testing with the
59:04understanding that it comes back
59:06with a higher percentage of cases,
59:08suspicious or positive.
59:09So you have to be
59:10careful with that. So.
59:16Off.
59:27Select.
59:34For the management based on the subject.
59:38Because sometimes those.
59:41Very in the nose part.
59:48We know.
59:56So grouping, excellent question.
59:57So that's why we wanted to simplify.
01:00:00So we don't recommend to the clinicians
01:00:03how to manage the US it's their choice.
01:00:07They know more about clinical history,
01:00:10clinical palpation findings
01:00:11or imaging findings.
01:00:13All we do is we report what we see.
01:00:16So previously they used to get
01:00:18very confused where we would say
01:00:20AUS where some of our bodies AUS
01:00:22rare internuclear inclusions.
01:00:24And by the way,
01:00:25the worst diagnosis you can give
01:00:27to a clinician is AUS with rare
01:00:30internuclear inclusions because.
01:00:32They are taught internuclear inclusion
01:00:34equals papillary thyroid carcinoma.
01:00:36So we should never do that.
01:00:38And these were the reasons that we
01:00:40wanted to tell our clinical colleagues
01:00:42that AUS is not a very complex,
01:00:45subjectively interpreted diagnostic category,
01:00:48that it could be very well defined.
01:00:53So nuclear is when you see nuclear, Ethiopia.
01:00:56Features of papillary but very
01:00:59folky present and everything else,
01:01:01including on Pacific suspending everything
01:01:03you will call other if you like.
01:01:06You can still put a descriptive
01:01:09diagnosis as a second line note.
01:01:12So your other question that you have was.
01:01:16How to manage more conservatively or
01:01:19lobectomy or otherwise it's a decision
01:01:23which is based on their discussion
01:01:26and but the reason I showed that
01:01:29slide where I wrote that now sometimes
01:01:31it simply follow up and they don't
01:01:33do anything that we have seen that
01:01:35they have started to realize that.
01:01:38If there's an AUS based on other
01:01:43reasons than nuclear tipiya,
01:01:45these patients will have a very low
01:01:48probability of having a malignant tumor.
01:01:51Has also shown by pharma testing that
01:01:55almost 80% of them turn out to be behind.
01:02:00Does that answer your question? Yeah, yeah.
01:02:05You have dialysis.
01:02:13Yeah, exactly. So ADA guidelines are
01:02:17actually built on the tester system
01:02:20which is which is wonderful news.
01:02:22And even now the 2022 I have seen a
01:02:28little bit of draft form and and again
01:02:32it's based on what we are suggesting
01:02:36in the in the forthcoming edition so.
01:02:40Which is great because we can very
01:02:43easily talk to each other using those
01:02:46potesta categories and know how the
01:02:49patient would be treated or managed.
01:02:52So let me just ask one more question, sure.
01:02:56Suspicious. That's great.
01:03:01Image means us.
01:03:06So they probably like Siri.
01:03:09And when you call those suspicious
01:03:12people, at the end of the day,
01:03:14you know the the risk of malignancy
01:03:17suspicious actually approaches 100%.
01:03:18So at the end of the day for when?
01:03:26So that's
01:03:28so 75% and Hopkins is 80%.
01:03:31Now I think in the new edition,
01:03:34a lot of people.
01:03:35We'll do the total thyroidectomy,
01:03:37but if I would,
01:03:39I'm calling something suspicious
01:03:41for proprietary carcinoma and I've
01:03:43convinced myself the the features
01:03:45are more than what I would see in
01:03:48a nifty and I don't want to call it
01:03:51flicker neoplasm only I put a statement
01:03:53recommendation for the first time.
01:03:56I tell people to put us a statement
01:03:59do lobectomy or consider lobectomy
01:04:02or whatever because although
01:04:04the official recommendation.
01:04:05This in the back we most surgeons
01:04:07they should go back with a
01:04:09centerfire 80% risk of agency
01:04:11and do as a second procedure of
01:04:14completing the higher directly.
01:04:16So they will do it if it's
01:04:19not a papillary carcinoma,
01:04:20the poor patient is way over treated.
01:04:23Are also people have asked me
01:04:25this question why do we even have
01:04:28the suspicious category when
01:04:29the risk of malignancy so high?
01:04:31And answer is very simple,
01:04:33we want to give the category of malignant
01:04:36a very high specificity approaching 100%.
01:04:40That's why we have suspicious or
01:04:42popularity or suspicious category in general.
01:04:45If we take it out then the accuracy
01:04:47or risk of malignancy of flat out
01:04:50malignant will drop down into
01:04:52lower 90s or even lower.
01:05:02That they have some.
01:05:09And the appointment category all give you.
01:05:17You mean the adenomatoid
01:05:19nodule case that I showed you?
01:05:22Yeah, so those cases would
01:05:24be extremely difficult not to
01:05:26call a US suspicious of PTC.
01:05:28The case which I shared in
01:05:31the morning unfortunately was
01:05:33called suspicious for papillary.
01:05:35Which you guys,
01:05:36a lot of you guys thought was
01:05:37suspicious for probably perfect, yeah.
01:05:42It's. Right. Thank you.
01:05:48So suspicious of February.
01:05:56Any institution that.
01:06:07Any.
01:06:13So difference is situtions
01:06:15handle it differently.
01:06:17So. At our institution,
01:06:20the trend is to do a a lobectomy.
01:06:24For suspicious of proprietary carcinoma.
01:06:27And if it turns turns out to be negative,
01:06:30we don't include it as a false positive.
01:06:32False negative.
01:06:35Or false positive, whichever way you look at.
01:06:37Because when we calculate the false positive,
01:06:40it's only flat out positive case.
01:06:44Some studies include both suspicious
01:06:46and and malignant when they
01:06:48calculate the false positive rate.
01:06:52If you like definitely each case
01:06:55of suspicious or popular you
01:06:57can recommend I'll actually.
01:06:59We sit sometime in tumor boards
01:07:01and and they don't mind if if
01:07:03if you tell them that you feel
01:07:05uncomfortable having the patient a
01:07:07total thyroidectomy because you're
01:07:09calling something suspicious papillary.
01:07:12But again,
01:07:13say that these days,
01:07:14with the with the NIFTY being
01:07:16such a such an issue,
01:07:17it's best that your diagnosis is suspicious
01:07:20or popular should be very carefully selected.
01:07:24Either it's papillary or subtle features.
01:07:28Focal features.
01:07:28Think of little and your passion.
01:07:31You can certainly write for follicular
01:07:33neoplasm differential diagnosis
01:07:34include NFP or a floccular variant
01:07:36of papillary thyroid carcinoma.
01:07:38We do it all the time.
01:07:40And molecular tests help and again
01:07:42answer to your question a lot of
01:07:45times suspicious for papillary now
01:07:46get molecular testing pharma which
01:07:49is based on negative predictive
01:07:51value 96% also has a second layer
01:07:54of testing called expression Atlas
01:07:57which is based on a 200 plus very
01:08:00specific positive mutations of thyroid
01:08:02cancers and and they can pick up.
01:08:05So if you call something as PC.
01:08:08Pharma is suspicious.
01:08:11GSC suspicion, the GC,
01:08:13Quincy classifier or pharma.
01:08:15They will add an expression Atlas
01:08:17over that and it's expression Atlas
01:08:20is negative with all the specific
01:08:22markers negative for papillary,
01:08:24other cancers they will not develop.
01:08:33Well, if there are no more questions,
01:08:34once again thank you so much.
01:08:36Have a great rest of the afternoon.
01:08:38I know it's a working day,
01:08:39so everyone is busy.