The Processes and Challenges of the WHO Blue Books

Name: The Processes and Challenges of the WHO Blue Books
Uploaded: 2021-10-26T21:00:02.26Z
Duration: 59 min 24 s
Description: October 21, 2021 Yale Pathology Grand Rounds Ian Alexander Cree, BMSc (Hons), MB, ChB, PhD, FRCPath, ALS

October 26, 2021

October 21, 2021

Yale Pathology Grand Rounds

Ian Alexander Cree, BMSc (Hons), MB, ChB, PhD, FRCPath, ALS

Information

ID: 7075
To Cite: DCA Citation Guide

Download Transcript

00:04Get started then.
00:40Hi everyone, welcome to today's grand round.
00:44Our speaker today is Dr. Ian Cray.
00:50Ian is a pathologist and an
00:54international civil servant.
00:56Is based at the International Agency
00:59for Research on Cancer in Lyon,
01:03where Ian is the head of humor
01:07classification responsible for all
01:09the blue books that we keep buying.
01:13And he is also head of evidence,
01:15synthesis and classification.
01:20He ended his undergraduate education at the
01:25University of Dundee and then his MBCHB,
01:30followed by a PhD in Immunology
01:33at the same place, and even then
01:38finished his pathology training.
01:40He was trained as a general
01:43pathologist at the same place,
01:45becoming a consultant
01:47pathologist in his alma mater.
01:51But since then,
01:52Ian has moved several times his
01:55previous positions held were senior
01:58lecturer and at the Institute of
02:03Ophthalmology and Ian mentioned that
02:06his major work was in ophthalmologic
02:09pathology and he got so excited
02:12with it that he was instrumentally
02:15involved with the first blue book
02:19on my pathology and his working.
02:22To bring out the second edition,
02:26or rather the 5th edition,
02:27what would be equivalent to the 5th edition?
02:31He went on to become consultant
02:34pathologist and a professor of
02:37histopathology at Queen Alexandra
02:39Hospital in Portsmouth and then
02:42Southampton in England and then
02:44became a professor of pathology at
02:47University of Warwick and a consultant
02:51pathologist at University Hospitals.
02:54In Coventry and Warwickshire,
02:56and just when he was getting ready
02:59to retire as the British system.
03:02And courageous their sin apologist to retire.
03:06He got an offer from Dublin to help
03:09them publish the new series of Blue
03:13books and become an international civil
03:17servant and of course moved to France.
03:22So just to give an idea that
03:25the series Blue Book,
03:27this is what they look like for those of
03:30us who are not in diagnostic pathology,
03:34and Ian has been involved with the 4th
03:37edition of the skin tumor Blue Book,
03:40the eye tumors and he is on the
03:44editorial board of the 5th editions
03:47of tumors of the digestive system.
03:50Breast humors soft tissue.
03:52In bone,
03:53female genital tract tumors
03:56and thoracic tumors,
03:58Ian's career has been built on
04:01translational path ologist,
04:03and on translational pathology,
04:05and he was involved in developing a
04:09number of molecular diagnostic methods,
04:13and he let you kiss early
04:16Cancer detection consortium.
04:18He has published more than 270.
04:22Papers and more than 150 of them
04:26were as first and last authors,
04:29and has been a coauthor in
04:32more than in about 8 books.
04:35So with that proof.
04:38Brief introduction.
04:40They will let Ian take the floor.
04:43Thank
04:44you very much. I'll start by saying
04:46thank you to to you for inviting me.
04:48It's a great pleasure to be here,
04:50at least virtually.
04:51I'm sorry I can't see the autumn
04:54colors around you and get to see
04:56and meet some of you properly,
04:58but maybe in the future. Who knows.
05:00I'm going to start by sharing my
05:02screen if I may, because I think
05:04that's the first order of business.
05:06Uhm, hopefully you can all see that.
05:09Uh, the.
05:12So this is the point where you have to
05:15shout if you can't because once I put
05:18the presentation on I can't see you.
05:20So I think
05:22we can see it
05:23good. Alright, let's go so.
05:27I was asked to talk about the.
05:30Processes and challenges.
05:31And this is not yet coming.
05:33Sorry though,
05:34it is of The Who blue books and.
05:40It's it's certainly had its challenges,
05:42and one of the challenges
05:44has been the process,
05:45so I think it's a fairly good title,
05:47actually. First of all,
05:48the declaration of interests.
05:50I am a pathologist has been said
05:51and based at the International
05:53Agency for Research on Cancer,
05:55which is a part of the
05:57World Health Organization.
05:58But the opinions I'm expressing
06:00our of course personal.
06:02I'm going to start with this.
06:03This is one of my favorite slides.
06:05Now I have to say it's a picture of Carl,
06:08Aeneas and Linnaeus described the
06:11natural world around us and we still
06:14use the classification that he put
06:16together and published in 1758.
06:19And if any of the blue books are still
06:21knocking around in 300 years time,
06:23I should be extremely pleased.
06:25But I'm afraid I won't be here to see it.
06:28The natural progression,
06:30though of all taxonomies,
06:32is that they evolved.
06:34And the natural our taxonomy of the
06:38natural world is also still evolving.
06:41This was a story from last Christmas.
06:44Nice Christmas story.
06:46Gentoo Penguins,
06:47which all look rather similar.
06:49There were actually two species,
06:52but now there are four,
06:53and that's on the basis of genetics.
06:55And that's exactly the same thing
06:56that's happening to a lot of tumors.
06:58Of course,
06:59where genetic investigation
07:01is proving very helpful.
07:03So taxonomy is the science of
07:05defining and naming groups of
07:08biological organisms on the
07:10basis of shared characteristics.
07:11And cancer classification is also based
07:15on the shared characteristics of cancers.
07:17That's currently mainly
07:19Histology and genetics,
07:20but it also includes things like radiology,
07:23cytology and a host of others.
07:27We take our remit for doing this
07:29from the 10th World Health Assembly,
07:32which in 1957 resolved to continue
07:35work on formulating international
07:37definitions of normal stature
07:39and statistical classification.
07:41The World Health Assembly is the governing
07:44body of the World Health Organization.
07:46It comprises 194 countries at the moment.
07:52I should say that I arc is a
07:55specialized agency of the World Health
07:57Organization and therefore it doesn't
07:59have the whole 194 to deal with.
08:02We've only got 28 and I think some of
08:04us feel that it's probably quite enough,
08:06but we would always like more.
08:09In the 4th edition, the WTO Bluebooks,
08:12which started in the 1960s, UM,
08:15became a synthesis of scientific evidence of
08:19illustrative cases and diagnostic criteria.
08:23They became indispensable to many
08:27pathologists because they gave you
08:30the international standards that
08:32were required to diagnose tumors.
08:35And I think that's where we owe
08:38them a great debt of gratitude.
08:40However, it must be said that
08:42not everybody was pleased.
08:44And those of you that have seen this slide
08:46before, and it has been around a bit.
08:48I got mine from that at Jabot and I'm very
08:51pleased to give her the credit for that.
08:55But here you can see a dinosaur.
08:57It's obviously a gynecological dinosaur,
08:59and it's eating a CNS blue book.
09:02I'm not sure they're actually suggesting
09:04that that guy, any pathologist,
09:05start reporting all the CNS pathologies.
09:07So don't worry too much.
09:09But certainly there were suggesting that
09:11these books were somewhat out of date.
09:14By the time they were published and they
09:16took about three years to publish each book.
09:19So that was one of the challenges the entire
09:224th edition took over 12 years to publish,
09:25so it was a mammoth task,
09:27and one that a smartly small
09:29number of people took on and did,
09:31and we owe them a great debt of gratitude,
09:34of course.
09:36However, we needed to innovate.
09:38We needed to change things,
09:39so there are a number of
09:41innovations for the 5th edition.
09:43And these really fall into three categories.
09:46The first was better governance.
09:48The entire 4th edition was actually
09:50led by just three pathologists,
09:53and that's probably nothing like enough.
09:57I certainly when I was appointed,
09:59didn't wish to try and lead it myself,
10:02and therefore I decided it would be a
10:04good idea to have an editorial board.
10:06And I borrowed from the National Institute
10:09for Clinical Excellence in the UK,
10:11which in its committees has both standing
10:14and expert members and the standing
10:16members are there to provide the
10:18continuity and the overall expertise,
10:20and they're usually
10:21fairly senior individuals.
10:22The experts are there to provide
10:25the hard experts opinion that's
10:28required to make decisions within
10:31health technology assessment.
10:33In the case of Nice or classification
10:36in this instance.
10:37And we try to do what we
10:39do based on evidence,
10:41and that now includes things
10:43like systematic reviews,
10:44which historically pathologists have
10:45not really been involved in very much.
10:49We are about evidence and not eminence.
10:53Expert opinion is still is still
10:57very relevant, but it's not.
11:00It's not of great significance.
11:03It's really important that we.
11:06Have also selection by informed
11:09Bibliometrics and we we we.
11:14We use both author and editor
11:17selection using this process,
11:19which allows us to look at what people
11:22have published and decide what is.
11:25What is is is most fair in
11:27in terms of who to appoint.
11:30It's not a simple process,
11:32it's something we have to do quite carefully,
11:34because clearly we don't want everybody
11:36to come from one country or another.
11:38We want to have a geographical spread as
11:40well as a spread of expertise and experience,
11:43and we try and do that using a system
11:45which we've actually developed called
11:47the Blue Book Expert Selection Tool,
11:50which is a form of cross.
11:54It uses cross referencing to identify
11:57papers published by authors and
12:00who they have Co published with,
12:02so it's a Co citation system.
12:05I'm not going to talk about that much more,
12:07actually,
12:07'cause it's it's quite complex.
12:10In terms of quality.
12:13Uhm?
12:14We've got a hierarchical classification
12:16using Linnane principles,
12:18and that gives us some tremendous scope
12:22to collapse and expand particular
12:26diagnostic categories quite easily.
12:29In terms of quality,
12:30I'll talk about that a little bit more later,
12:32and we've certainly upped the quality
12:34of images we tried to ensure that
12:37references are up to date and relevant,
12:40and there's quite a bit more.
12:43We get our epidemiology from epidemiologists,
12:46which is a bit of a novel turn,
12:48but the epidemiologists in arc and
12:50there are a lot of them were fairly
12:53scathing about the pathologist's
12:54ability to do this, so we said,
12:56well, why don't you do it?
12:58And luckily, they said yes.
13:00We try to incorporate new
13:02information whenever we can,
13:04and we harmonize topics across series,
13:06particularly neuroendocrine neoplasms,
13:09humita,
13:09lymphoid disease and soft tissue tumors.
13:16We've improved the speed with which
13:18we can produce the books by just
13:21taking a process management approach.
13:23It's lean management, really.
13:26So we've tried to improve things
13:29at every stage by setting deadlines
13:31and trying to meet them and also by
13:35improving the way in which we handle
13:38matters in an automated fashion as far
13:41as possible within the software that
13:44we used to produce the whole thing.
13:46We have a website that allows easier
13:49access to references to whole slide
13:51images and it now allows notes and
13:53it also allows feedback so that
13:56then allows those that are part
13:58of the community that use and buy
14:00the books and the website too.
14:05Come back to us in a much easier way
14:07and a much more organized way too,
14:09which is probably better.
14:11So the 5th edition books look
14:13a little bit different.
14:13The 4th edition,
14:15they've got a slightly modernized cover.
14:17There are slightly lighter blue,
14:18so you can tell which is which on the shelf,
14:21and there's quite a lot of thought
14:23gone into the design of both
14:24the cover and the spine.
14:26We've kept the nine photographs in the
14:28front 'cause that makes them instantly
14:30recognizable as a WHO blue book.
14:33And we've gone for a two column
14:35format inside.
14:36So the photographs,
14:37which were rather small in the previous
14:40edition in a three column format
14:41and now on the whole much larger,
14:44and that certainly helped.
14:46We've incorporated clinical photographs.
14:48We've incorporated epidemiological
14:50graphs as well. Uh, diagrams.
14:53So there's a wealth of information in these.
14:56Volumes,
14:57but I should say at the outset
14:59they are not textbooks.
15:01This is a taxonomy would be describing
15:04the criteria for each tumor type,
15:06which results in that classification.
15:08We're not trying to tell you
15:10how to diagnose them,
15:11that's something that a textbook would do.
15:14In terms of who?
15:16Therefore,
15:17the other thing is that they're
15:19not just for pathologists,
15:20this is the taxonomy of cancer,
15:22and just as a gardener or a.
15:27A farmer will be interested
15:29in the plants on their land.
15:31People with different ideas of of
15:33what they want are quite capable
15:36of looking at the same taxonomy
15:38and getting something from it.
15:40The doublet show blue books then
15:42provide a definitive evidence based
15:44classification of all cancer types to
15:46enable diagnosis and research worldwide.
15:51And our diagnosis of cancers
15:54and other tumors underpin
15:56individual patient treatments.
15:58Of course, that's why we do it,
16:00but they also underpin research
16:02into all aspects of cancer,
16:04causation, prevention and therapy.
16:06And then they also underpin education.
16:09So it really is the basis of
16:10an awful lot of what we do.
16:14In terms of the classification turns,
16:16we use usually A5 level
16:20classification with subtypes.
16:22As a further level.
16:24So we start off, usually with sites,
16:27for instance stomach.
16:29We then have a category which is
16:32often lineages based epithelial plasm,
16:34something like that and then a
16:37family or class which in this
16:39example I've given as abnormal other
16:42pre malignant neoplastic lesions.
16:44We might separate those and some other
16:46volumes and then a type would be adenoma.
16:49So you'd expect to see gastric
16:52adenoma in the diagnosis perhaps.
16:55And then the subtype, UM,
16:58would be public land in this example.
17:01Variant is not a word.
17:03We try to use anymore within the blue books.
17:07In terms of the Histology or
17:09what is now called the subtype.
17:12The reason for that is partly because
17:14it's been somewhat hijacked by others.
17:16It's used variably by geneticists and
17:19by others to describe variations that
17:21may or may not be part of the classification.
17:24So we've tried to avoid it where possible
17:28and deal with it when we absolutely have to.
17:32We haven't, though I should say,
17:34switched from calling mutations as you
17:36may very well know them sequence variants,
17:40which is what most geneticists would do now.
17:43In terms of staging grade,
17:45those are dealt with separately
17:48within the books.
17:49Honestly, the early sort of
17:51points that we realized was that.
17:53Your view of the classification really
17:55depended on what you were doing.
17:58So if you were a geneticist or a histo,
18:00pathologist or radiologist,
18:04you looked at things very differently.
18:06And that's a multidimensional problem
18:09and it's very hard to put multiple
18:11dimensions into a two dimensional book.
18:14So we decided to make it slightly easier
18:17on ourselves by having a database.
18:19So the classification is now a database.
18:22It's a relational database and it means
18:24that we can now use the headings for
18:28each of the tables to develop the two
18:31dimensional layout you see in the books,
18:34and that's why you will see
18:37things like etiology unknown.
18:40It's not because we didn't.
18:43Yeah,
18:44we didn't leave etiology Alex because
18:46there was nothing to put in there.
18:48We actually have to say something for
18:51each of these headings in the book.
18:53We don't, uh.
18:57Make any distinction underneath these
18:59headings so that the open bullet points are.
19:03You can be used or not used as as
19:05people wish and you'll see them
19:07in some of the larger sections,
19:09particularly where it breaks up the text,
19:11but you won't otherwise see them at all.
19:16So we start off with the definition.
19:17The definition now has to be less than
19:20100 words and it has to define what
19:23something is and not what it's not.
19:25And it has to be usable by
19:28people outside pathology.
19:30Uhm, because it is used on its own,
19:32it's used for instance,
19:34with the ICD 11 codes.
19:36I won't go into coding 'cause
19:37that is pretty specialist,
19:39but suffice it to say that
19:41that we have the International
19:43Classification of Diseases on koleji.
19:46Uhm, which is used largely
19:48by cancer registries,
19:49and we have the International
19:51Classification of Diseases Version 11,
19:53which is the new one that's taking
19:56over from ICD 10 at the moment.
19:58And it's still really under development.
20:02We give related terminology and
20:04we divide that into two types,
20:07acceptable and not recommended.
20:09And that's because we don't want
20:12to stop people doing things.
20:14It's not our role,
20:15but it's certainly necessary to
20:17say what is a good idea and what
20:20perhaps isn't such a good idea.
20:22We list subtypes and we described
20:24them under whichever category they
20:26all categorization they come into.
20:29So if it's a histological
20:30subtype comes under Histology.
20:31If not,
20:32it might be a genetic one described
20:34under the molecular pathology.
20:38The clinical features and radiology come
20:40next and we put them together because
20:42they seem to fit fairly well that way,
20:44at least in in the first book we did,
20:47which was in digester.
20:49Uhm, epidemiology.
20:50I've talked about already and etiology.
20:53We covered the causes and this
20:56predisposing factors going here so
20:58that the predispositions genetic
21:00predisposition would go in there.
21:03In terms of pathogenesis,
21:05that is largely these days molecular,
21:08but we do try to make sure that we cover
21:11other other parts of that as well.
21:13A macroscopic appearance in
21:15histopathology probably needs no
21:17introduction to this audience,
21:19but it has all the usual things
21:21you'd expect under there.
21:22We try to have differential
21:24diagnosis at the end,
21:25which is very helpful to to
21:28many pathologists I know.
21:30In terms of cytology,
21:32that is now being dealt with,
21:34in addition to the blue books as
21:36a series of reporting systems,
21:38foresighted pathologists and we've
21:40teamed up with the International
21:42Academy of Cytology to produce those,
21:44and I'll show you a mock up of the
21:47cover later in terms of diagnostic
21:50molecular pathology that is restricted
21:52to what you would do in a patient.
21:54So if you wouldn't do another patient,
21:56it won't appear there.
21:57It'll appear on the pathogenesis.
21:59One of the innovations that we have
22:03introduced is the addition of essential
22:06and desirable diagnostic criteria.
22:09And the reason we've done that is
22:12because in the previous 4th edition
22:14it was sometimes quite hard to
22:16find out which were the essential
22:19features that you needed to see in
22:21order to make a diagnosis or to
22:23classify tumor within a certain range.
22:26And that's something that we felt was.
22:29It was a good idea to actually write down,
22:31so that's what we've tried to do.
22:32It's quite a challenging experience,
22:34and some of you I know in this
22:36room have had it.
22:38We use the UICC staging system
22:41because we are international.
22:42This is the international staging system.
22:45But UICC and AJ CC work
22:47very closely together,
22:49so if you're using the AJC,
22:51it's the same thing.
22:53In terms of prognosis and prediction.
22:57We talk about prognostic factors
22:59and predictive biomarkers in some
23:01detail in certain tumor types,
23:03where that's really important,
23:04and that's an increasing number.
23:05Of course,
23:06and then we're developing links
23:08to other resources which will
23:10mainly go on the website,
23:11and that's particularly to the international
23:14collaboration for cancer reporting.
23:16CAP is one of the founding
23:18members of that organization,
23:20as is the Royal College of
23:22Pathologists and many others,
23:25and.
23:26It produces the structured reports which
23:29depend very heavily on the blue books
23:33and the staging resource is produced by UIC.
23:37In terms of numbers,
23:38we have roughly 200 people
23:40on the editorial board,
23:43between the standing and the expert members,
23:45or will have by the end of the series.
23:48The authors are really need to
23:50update that figure because we've
23:51already hit two and a half thousand.
23:53Uhm, so it's a it's a very big project,
23:57this one and it has a lot of
23:59people involved in it and we're
24:00extremely grateful to all of them.
24:02It's a it's a mammoth task and it has
24:06enormous impact for patients particularly.
24:10In terms of subscribers,
24:13we think we have about 60,000.
24:15They are mainly pathologists,
24:17but not entirely probably
24:19around 8590% pathologists.
24:22And if you're anything like me
24:24when you reach for your blue book,
24:25it's not there because somebody borrowed it.
24:28And therefore we think we've got a lot more
24:31users than we have actual subscribers.
24:33Uh, the whole team is the whole thing,
24:37is run by a small team of people in
24:40Leon and and there are twelve of us
24:42and it's it's great to have them there.
24:45It is possible to join us.
24:47We have a number of fellows that joined us
24:49from time to time and visiting scientists.
24:52So if you've got the opportunity
24:54and you'd like to do it,
24:56we can fit a couple of people in,
24:57usually into the office space we have.
25:02And it would be lovely to welcome
25:04people from Yale if we could.
25:05In terms of what we've done already,
25:08well, we've published digestive breast,
25:10soft tissue and bone female
25:12genital and thoracic tumors.
25:14The central nervous system volume
25:16is about to be stitched together in
25:19from the individual chapter proofs
25:21and it will go to the printers
25:23on the 12th of November.
25:25So it will appear on the website,
25:27probably towards the end of
25:29this month as well.
25:31In terms of the pediatric tumors volume,
25:34that is massive.
25:36It's the equivalent of two books,
25:38and it needs to be coordinated
25:40with quite a lot of the others,
25:42so it's proving quite a tough one to do.
25:44And and we've got the draft,
25:46but we haven't met yet.
25:48Managed as yet to get it through technical
25:51editing, so it is a bit delayed,
25:52but it is on its way.
25:55The urogenital male genital
25:57tumors is also impressed.
25:59That is with the technical editor
26:01at the moment and it should come
26:03out relatively early next year.
26:05I hope within Q1 or possibly Q2.
26:09In terms of the head and neck and the
26:11endocrine near endocrine and endocrine,
26:13they are being edited at the moment's ready
26:17for technical editing early next year,
26:19so they should be out sometime
26:22over the summer.
26:23And then the humours of lymphoid
26:25tumors volume is being written.
26:26Uh,
26:27and then we're about to start
26:29the skin and the I.
26:30We set our timescale for doing things
26:33at the beginning of the process so
26:35that that we could get the whole
26:37thing done within the five to six
26:39years that we felt was optimal
26:41and and that was backed up by the
26:44findings of a large survey that I'm
26:46sure some of you participated in.
26:48And again, thank you for doing that.
26:50And we're going to finish the whole
26:52series of the 14th volume on genetic
26:54tumor syndromes because that's
26:56becoming increasingly important that
26:57we felt we could not ignore that.
27:01The website now has a new look,
27:05and if you haven't seen it,
27:06I'd recommend getting onto it.
27:08If you can, you can preview it before
27:10you have to pay a subscription for it.
27:13It is a subscription website and we
27:15do sell the books and the reason for
27:18that is that this is a not for profit
27:20exercise that is completely funded by
27:23sales of the books and and the website.
27:27Now, yes, we could apply for grants and
27:30we could try and find external funding.
27:33But then that external funding would
27:35have a say in how we did things,
27:37and that's not something we want
27:39at the moment.
27:40This is essentially in the
27:42hands of its subscribers,
27:44so that's 60,000 people and it's
27:48very hard to make it anything other
27:50than what it is as a definitive
27:52classification on that basis.
27:57The actual write up for each tumor
27:59looks very similar to the book.
28:00It's just a single column across a text,
28:02but it's the same text, the blue numbers
28:05there are actually pub Med ID's,
28:07so when you click on one of those,
28:09a hyper link takes you straight to pub
28:11Med to the reference and to the evidence
28:14we've quoted for that particular statement.
28:16We also have whole slide images
28:18and in the breast, but we have
28:21whole slide images for every tumor.
28:23And when you click on it,
28:25you get a legend just as you
28:27would for anything else,
28:28and then you get into the slide and then
28:31you can go to very high power view.
28:33So I was asked to talk about
28:36the challenges and.
28:38I expected some of them and the
28:41expected ones included the fact that
28:43we needed to produce these volumes
28:45faster without compromising quality.
28:47We knew that was going to be
28:49something that was important.
28:50We also knew that whole genome
28:52sequencing was coming into mainstream
28:54diagnostic practice for some things,
28:56and XM sequencing for others.
28:59We knew that there was a lot of
29:01genetics that we had to get in.
29:03We kind of thought artificial
29:04intelligence probably a little bit
29:06too early for the 5th edition,
29:07but it has actually come into
29:09one or two points already.
29:11And it's certainly set to do more.
29:14We did think that perhaps radiologists
29:16were trying to take over from
29:18pathology in terms of some diagnostic
29:20entities that hasn't really happened,
29:22but we're very pleased to say that we
29:24do have a radiology Advisory Board
29:26that gives us an enormous amount of
29:29support in making sure the radiology
29:31within the books is as good as it can be.
29:34The radiologists actually
29:35don't have anything like it,
29:36so they're really quite interested
29:37in what we're doing.
29:40There was some unexpected stuff too,
29:41of course. Look, the confusing terminology
29:45within pathology is something that
29:47most of us have lived with and dealt
29:49with for most of our working lives,
29:51but we did have some really.
29:54We do have some really good ones and
29:55we are trying to reduce the confusion
29:58wherever possible because it does
29:59impact on patients occasionally.
30:01The best one we came across was,
30:03I think, in the digestive volume
30:04and it was this one inflammatory
30:07pseudo tumor like follicular stroke
30:09fibroblastic dendritic cell tumor.
30:11And all of us in the room kind of looked
30:13at one another and said, what is it?
30:15And and then looked at John Chan,
30:17'cause he's always got the answer.
30:18And and on very kindly turn around and say,
30:21well, it's an EBV positive
30:23inflammatory dendritic cell sarcoma.
30:25So we said, well, why did we call it that?
30:28And the answer was that we did,
30:30so it's one of the very few that we've
30:33actually taken essentially outside
30:36the system and made a change because
30:38we felt the name is just impossible.
30:41Uhm, we have a problem with Mitoses
30:45and the major problem there.
30:47I'll show you in a moment,
30:48but it's to do with standardized
30:51international units.
30:52We had committed very early on
30:55to use HGVs notation,
30:57so the huger notation for genes,
31:00mutations, fusions, rearrangements,
31:02alterations and sequence variants.
31:05And we said that we would do this.
31:08It proved to be quite a challenge,
31:10because although molecular pathology
31:12has become very common in some settings,
31:15in others, it is more challenging.
31:20It turns out that that we have some
31:22problems in terms of counting as well,
31:25particularly bases and histones,
31:27and we've had to publish a paper on
31:30that to produce a notation that works.
31:34Uhm,
31:34and then of course we have this
31:36problem of whether we call things
31:38variants or subtypes and what's
31:40histological pattern when it's at home.
31:42So there have been a few things and
31:45I'll go through a few of them just now.
31:48One of the first things though,
31:50was to think about evidence based pathology.
31:55When in the 4th edition, the books
31:58are essentially a consensus document,
32:00a consensus statement of experts and
32:04about 30 to 40 experts involved in the
32:08meeting that would produce the consensus
32:10based on what authors had written.
32:12And the question really in our minds was
32:14whether this was eminence or evidence.
32:16And the answer is it probably is evidence.
32:19It certainly is evidenced because
32:21those experts are expert for reason.
32:24Uhm, but here you've got someone
32:26who is extremely eminent.
32:28The future King of England.
32:30But I'm not convinced that you
32:31want him staring at a microscope
32:33looking at your biopsy.
32:34Or mine for that matter.
32:37In terms of of confusing terminology.
32:40I've already mentioned one.
32:42I'm going to just select one here
32:44and it's pseudo tumor.
32:45That's about halfway down.
32:46It's a term used to refer to
32:49any number of pathologists,
32:50both benign and malignant,
32:52that may produce a mass.
32:54It's imprecise,
32:55it's not a good term,
32:57and we've tried to extract it from
32:59the books and over nearly succeeded.
33:01There is one in one of the books.
33:03For those of you that want to
33:05go looking for it.
33:06But there are other things here
33:07that we've known for a long time.
33:09Things like carcinoid,
33:11which is potentially imprecise as well,
33:14but is still very much used clinically in
33:17the lung field and impossible to change.
33:20Certainly in this edition.
33:24In terms of assessment of mitoses,
33:26this was something that I first
33:28came across nearly 40 years ago.
33:30And I was really surprised to
33:31have to publish on it again.
33:33I have to say. Uhm?
33:37First of all, what do you call it?
33:39Well, it's a mitotic count.
33:40When you do it on a microscope,
33:42nearly always.
33:42And it's really a density measurement.
33:45So you are expressing the number term
33:48of mitoses per millimeter squared.
33:50Not pleased high power field
33:52as I'll show you in a moment.
33:55The mitotic index would be the
33:57number of mitosis is expressed
33:58as a percentage of the number
34:01of neoplastic cells present,
34:02and you'd have to count hundreds
34:04of cells in order to do it.
34:06It's definitely one for the artificial
34:08intelligence and and not for the human,
34:10I would suggest.
34:12Mitotic rate is actually the
34:14rate at which cells are entering
34:16mitotic phase of the cell cycle,
34:19expressed as a percentage
34:20of cells counted per hour.
34:22So it has a time in it.
34:24And that's not something you can
34:25really do in a form and fixed.
34:28Specimen.
34:28So in practice,
34:30pathologists use mitotic count but we
34:33often get the name and the numbers wrong.
34:36Why do we get the numbers wrong?
34:38Well,
34:39historically,
34:39we've tended to use high power fields,
34:42ignoring the fact that between microscopes,
34:45high power fields can differ by up to
34:48six times even at the same magnification.
34:51And the switch to digital
34:52pathology is actually not helped.
34:53It's paint things worse because
34:55the 400 times fields in digital
34:58systems are rectangular and
35:00they also vary in area and they
35:02can vary quite a lot in fact.
35:05So the potential for error is quite
35:07considerable and at times that could
35:09affect patient care or even diagnosis.
35:11It's solved really by the use of
35:15standardized international SI units,
35:16so size does matter and it's millimeters.
35:20So we recommend counting features
35:22like mitoses per millimeter
35:24squared with an indication of the
35:26area to be counted and the method
35:28used which can be hot spot or
35:30average to obtain the results.
35:33These kind of graticule's on slides
35:35are actually available very freely.
35:37You can get them from Amazon
35:39for about $13.00 each.
35:39If you haven't got one and like one
35:42there should be one knocking around
35:43the department somewhere I'm sure,
35:45or possibly several and.
35:48The Convention is to use a 40 times
35:52objective, but if you use a 10
35:54times IPS or a 12 1/2 times IPS,
35:56of course you'll get something
35:58different and it's pile squared.
36:00Calculate the area.
36:01You only have to do it once
36:03for your microscope.
36:04I always had it on a sticky label on
36:05my microscope when I was reporting
36:07and it was something that I
36:10referred to virtually all the time.
36:13In terms of counting mitoses as well,
36:16it's important to know whether you're
36:18doing an average or random count,
36:20which may not be all that random
36:22unless you use a random number
36:24generator to do it properly,
36:26or a hotspot method and the hotspot
36:29North method is what is generally used,
36:32so these are contiguous.
36:36If. Microscope fields around a hot
36:39spot where you found a lot of mitoses.
36:43It's a little bit hit and miss,
36:45but it's better than most other
36:47things and it's been shown to
36:49be in various publications.
36:53There are quite a few examples
36:55here of of tumors, and this is only
36:57looking at the first few books,
36:58so it's by no means comprehensive,
36:59but it's just a few examples of things
37:01where there is a problem and reassessment
37:04of mitoses is probably worth while.
37:07The other reason why it's worthwhile to
37:09reassess mitoses in in tumors is because
37:12it is very often used prognostically.
37:15And prognosis changes depending on what
37:17the treatment is and if the treatment
37:19of the tumor has changed overtime,
37:21then what was done previously in terms
37:25of prognosis is a bit of a problem.
37:28I used to teach a lot of ophthalmologists
37:31about retinoblastoma and I would
37:33start by telling them all the
37:34things that were bad prognostic
37:36factors for retinoblastoma patients,
37:38and then say at the end of the lecture right,
37:41you can ignore all that once
37:42you pass the exam,
37:43because actually all these
37:45patients get chemotherapy and
37:47they all virtually all get cured.
37:50In fact,
37:50we had one death in eight years
37:52when I was at Morefield,
37:53so it is something that has to be looked at.
37:58In terms of treatment as well as pathology.
38:03We have a little bit of a
38:05challenge with statistics as well,
38:06so most biological data is skewed
38:08and if you use a mean you've got an
38:11error there which you don't have.
38:13If you use a medium,
38:14so the median is preferred by
38:17statisticians with a range because
38:19it's not affected by skewed data.
38:21In terms of genomics and most of you,
38:24I'm sure we will know this, but.
38:28B RAF V 600 E UM is written
38:30in all sorts of ways.
38:32This is the correct way to write it,
38:34and it's really important that that's done.
38:37If it's done with sequencing and
38:39if it's done using an antibody,
38:40then it's still important to say
38:43B RAF P dot V600E if you can,
38:46simply because that then tells everyone
38:49that you're looking at the protein.
38:51And the beer after course of its
38:55approaching should not be in italics.
38:57We very often, though I don't know,
38:59and I'm sure some of you have you
39:01sit in the clinical meeting and
39:03someone asked the B raft positive or
39:05negative on the Melanoma completely
39:06missing the point about mutation.
39:10In terms of histones,
39:11the problem is that the methionine,
39:12the little green one here and gets chopped
39:16off during production of the protein.
39:19And what that means is that the
39:22antibodies are all named according
39:24to the numbering of the protein.
39:26Once the methionine has been taken out
39:29and not what you get on a sequencer,
39:32which will give you a 28,
39:34not a 27 or 35, not a 34.
39:37So we've suggested that instead,
39:39what one should do is put the
39:43histological number if you like in
39:46brackets and the genetics number in
39:48genetics derived number in as you
39:52would normally in a in the system,
39:55and for some reason I've got a
39:57little hyphen appeared there,
39:58which wasn't supposed to be there.
39:59Apologies for that. Uhm?
40:02The fusion genes.
40:03This is literally hot off the press.
40:06This hasn't even got a payment yet,
40:08but it's just come out in leukemia
40:10and it suggests that rather
40:12than the previous notation,
40:13which was very confusing for fusion genes,
40:17we should use a double colon instead
40:20as the current system of using a hyphen
40:25or a forward slash is actually used
40:29in lots of other settings as well, so.
40:33The clever people Hugo came up with,
40:35and Elizabeth Bruford is is one of them.
40:39Came up with this idea of using two colons,
40:41and I think it's going to take.
40:43We are implementing that within the blue box.
40:47If you're interested in genetics are awful,
40:49lot of resource is available.
40:51I would point out variant
40:52validated down here,
40:53which is actually quite a
40:55useful one and we do as well.
40:58Obviously use some of the other things here.
41:01Gene names and gene cards I
41:04use quite a lot as well.
41:06I want to spend the last sort of 10
41:08minutes or so talking about research needs.
41:12Uhm, because we have lots of evidence gaps,
41:14we know they're there.
41:15If you look for the word unknown within
41:18virtually any of the blue books,
41:20you'll find 150.
41:22Also instances of it.
41:24And those are the things we
41:25know we don't know.
41:29In terms of of evidence gaps then.
41:33We have all sorts of things that
41:35have come out of the blue books,
41:36so we can actually say right?
41:38There's a need for more genetic
41:40studies of neuroendocrine tumors.
41:41There's a need for
41:43computational studies of Chi,
41:4467 perhaps of mitotic count.
41:47And we need to know what there is.
41:50So this was a this.
41:52This little list came out of a
41:54meeting on neuroendocrine tumors,
41:55which led to the current classification
41:58of new endocrine tumors,
42:00which has been applied across the series.
42:08I'm a Scotsman and I'm sorry I've got
42:09to give you a little bit of burns,
42:11but there is a translation at the bottom
42:12for those of you that are having a problem.
42:14UM, facts are chiels at winner
42:16Ding a downer be disputed. Uhm?
42:19It's essentially an older version of facts
42:23that are things that you can't dispute.
42:29And. As Donald Rumsfeld put it,
42:32we have known knowns.
42:33We have things that we know we know.
42:36They are straightforward.
42:38They're not under any.
42:41Control the C in pathology or anywhere else.
42:44And there are plenty of those
42:45in the blue books,
42:46and it's sometimes hard to
42:48find a reference for them,
42:49because actually we know that true.
42:52Then there were the known unknowns.
42:55There are things we know we don't know,
42:58and those are the unknowns that I've
43:01talked about in the books so far.
43:03But there are also undoubtedly a lot of
43:06unknown knowns things that we've forgotten,
43:08and a lot of that these days has
43:10come from things like electron
43:12microscopy and ultrastructure,
43:13which is not very fashionable and
43:15and it's quite surprising how little
43:17of it gets into the blue box,
43:19but I have pleased to say that in
43:22the endocrine volume you'll find
43:24that does happen because it's
43:26still there still important.
43:29And then we have the unknown unknowns.
43:30The things we don't know.
43:31We don't know,
43:33and that's something that
43:36really underpins research,
43:38so by looking at tumors carefully
43:41by looking at good well call
43:44it well put together series.
43:47And doing properly constructed
43:49cohort studies with.
43:51Protocols before you start the study,
43:54we can certainly get a lot further
43:56than we do. I think at the moment.
44:00And in terms of improving research quality.
44:03Why should it have lower standards
44:05when it affects gonna potentially
44:06affect huge numbers of patients than
44:09we would accept in clinical practice?
44:11So in clinical practice you would not
44:14accept a sample into your laboratory
44:16that didn't have 4 identifiers.
44:18But you would quite happily accept that into
44:21a research lab with just one in many cases.
44:25Drug approvals require two independent
44:28confirmatory phase three clinical studies.
44:30And what does pathology need?
44:32Well. I think it needs levels of
44:35evidence the same as anything else.
44:37We struggle with systematic reviews
44:39and pathology because they have to
44:42be systematic reviews of high level
44:44evidence and we struggle with getting
44:47hold of independent prospective studies.
44:49Of independent cohort studies
44:51of sufficient size.
44:52Sample size calculations are
44:54rarely stated in pathology papers.
44:56They need to be.
44:58Uhm and withdrawn often to case reports,
45:01small cohort studies or expert opinion.
45:05And that's still evidence.
45:06But it's not very high level evidence.
45:12So to try and answer this,
45:14having put together a process for
45:17production of the classification of tumors,
45:19we thought it was important to try and
45:22support the other side of the coin through
45:25setting up the international collaboration
45:28for cancer classification research.
45:30Uh, what that's about is,
45:32UM, evidence based.
45:34Science about best practice,
45:37guidance about standards and accreditation,
45:41and about information systems as
45:42well to try and reduce the vast
45:45amount of data produced every year
45:47to something that's manageable enough
45:49to pull through into a blue book.
45:52And at the same time we can feedback
45:55the priorities and gaps that we
45:58identify during the bluebooks process
46:00into the sound of that community.
46:03And this seems to have hit a bit
46:04of a chord with a lot of people,
46:06and it is a collaboration of institutions,
46:08not individuals I should add.
46:10So if you would like to join us,
46:13I'm sure that would be a great,
46:15gratefully received a lot.
46:17R and we're very grateful to those
46:20that have helped to set this up.
46:22I'm just going to show you one project from
46:24it and this is called the WCT evidence map.
46:29And it's so WTTW classification of tumors,
46:34obviously,
46:35and what this allows us to do is to
46:39have a tool which allows us to examine
46:42the blue books to identify what the
46:45evidence is and where it's lacking.
46:48This is a stylistic
46:50interpretations of maps you get,
46:52and that's what a real one looks like,
46:53and I don't expect you to read it,
46:55but you can see the blobs that are red,
46:57that I'm afraid is the low level
46:59evidence and the little green ones are
47:01the ones you really want to see Blues.
47:03OK, but green is high level evidence,
47:05and in epidemiology there's
47:07remarkably little for some tumors,
47:09but of course it does depend
47:11on how common the tumor is.
47:12It depends on how much funding
47:14has gone into researching it,
47:16and all sorts of other things,
47:17so it's not surprising.
47:18But we got this,
47:20but I think it's important information to
47:22know so that we know what we don't know.
47:28I'm nearly finished,
47:29but to just break come back to the
47:32site of pathology reporting system
47:34and talk about the mock up for The Who
47:37psychopathology reporting system volumes,
47:40which will be coming out next year.
47:42And as you can see on the front,
47:44we've got the World Health Organization IIRC,
47:47and we've also got the
47:50International Academy of Cytology.
47:52It will look a little bit different
47:54and it will have this. Uh.
47:56I think it's it's probably fuchsia,
47:59but we have a layout editor who
48:02is far better at color than I am.
48:05And she thought that was probably quite
48:07a good idea to have something that
48:09defined it meant that you didn't confuse
48:12it with a blue book on the bookshelf.
48:15We are still very much interested
48:17in working on AI,
48:19and one of the things that is fascinating
48:22to me and it's it's part of some
48:25work I started when I was in Warwick.
48:27Is that there is?
48:29Clearly,
48:30a relationship between the genetics
48:31of tumors and their histopathology.
48:33All there should be.
48:34Because there's a relationship,
48:36but doing the genetics of tumors
48:37and their behavior,
48:38so we should be able to see that
48:40reflected in this to pathology.
48:42It's actually been quite a hard
48:43thing to demonstrate,
48:44but here in colorectal cancer using TCGA
48:47data we've been able to do just that,
48:51and the rock the rock curves here.
48:53The receiver operator characteristic
48:55curves and not too bad the area
48:59under them is about .89 best,
49:02which is almost at the point
49:03where you start thinking about
49:05putting it into practice,
49:07so it may be possible to
49:09identify some of the genetics.
49:12That we need to treat patients
49:14with from artificial intelligence
49:17based assessment of histopathology.
49:20And the other thing it tells us is
49:22which are actually the important
49:24features that we talk about when we
49:27look down a microscope for defining the
49:29characteristics of the patients tumor.
49:34So I'll finish by talking
49:36about how how you can help.
49:38First of all, if you can provide
49:40the evidence by doing research,
49:42that's a big help.
49:44Uhm, and and don't think that when you
49:46published the paper that's the end of it.
49:48It's a question of taking that research and.
49:52Pushing it in the direction where it will
49:55get into practice if at all possible.
49:57And then evaluate the evidence.
49:59Systematic review learn how to
50:00do systematic review properly,
50:01and it's not simple.
50:03You have to learn how to write protocols
50:06for it and assess evidence for bias,
50:11particularly.
50:12And it is quite a challenge to look
50:14through large numbers of papers.
50:17And then if you buy the books or the website
50:20you are actually funding the evidence.
50:21So thank you very much indeed for doing that,
50:24because that keeps the whole thing going.
50:26And finally, let us know what you think.
50:29You can give us feedback on the website.
50:31You can certainly email us a lot of people.
50:34Do I get an awful lot of emails and
50:36we're always open to receiving cases
50:39which are better representative
50:40cases than those in the books.
50:43And if course, if you find any errors,
50:46do tell us because that's
50:48the way we get them fixed.
50:50So in conclusion,
50:51there is a need for all cancer
50:53diagnosticians to contribute to
50:55research to gather the evidence
50:57a patient needs and to evaluate
50:59that evidence for use in practice.
51:02Our diagnosis underpinned the
51:04management of individual patients
51:06Cancer Research and epidemiology.
51:08It's pretty important stuff.
51:11In terms of implementation
51:13of academic research,
51:14pathology is almost unique.
51:18In being able to do that itself
51:21through the WTO classification of
51:23tumors which provide the international
51:25standards for knowing motors,
51:26you can't do that with drugs and you
51:29can't do it with a lot of other things.
51:31Which gulf governments really
51:34are the arbiters of?
51:36So we have a joint responsibility
51:39to ensure the accuracy of the
51:41classification and to fill gaps
51:44and knowledge wherever possible.
51:46So thank you for your attention.
51:49This is the tower which I
51:51are currently inhabits.
51:52It isn't in great condition and the
51:55French government is very kindly building
51:58us a new Center for the International
52:01Agency for Research on cancer,
52:03which will really be a focus
52:05for international Cancer
52:06Research around the world.
52:08And a nice view of of of Lyon in the summer.
52:12And thanks to my team in Leon,
52:15which has worked so hard,
52:18particularly on the 5th edition.
52:20And I'm also grateful to two
52:22distinguished visitors, Valerie White,
52:24from Vancouver and Delani Luca Hattie,
52:28from Colombo, Sri Lanka.
52:31Thank you very much.
52:34Thank you so much Ian,
52:37that it was really a pleasure.
52:40To understand what goes behind the
52:43scenes in producing these beautiful,
52:46extremely affordable and
52:48extremely user-friendly books,
52:50I do have a question.
52:52Since you mentioned how difficult
52:55it is to gather evidence.
52:58So in pathology,
52:59most of the work we produce is
53:04retrospective and observational.
53:06And it's really getting harder to get.
53:11Follow up on many of these patients,
53:13so is there any future for
53:17retrospective and observational data?
53:20Yes, I think I think there is an I
53:24mean the the. The there's certainly,
53:26you know it's very easy when you
53:28look at levels of evidence and you
53:30say Oh dear opinions worth nothing.
53:32Therefore, the review I published last month,
53:33I might as well not have bothered.
53:35I did a case report last year
53:37with one of my juniors.
53:38I shouldn't bother doing that either.
53:39That's not the case.
53:41Actually, my sixth case report my
53:446th postmortem autopsy when I was
53:47training led to the withdrawal
53:49of a drug from the market.
53:50It was dangerous.
53:52So Uhm, case reports matter.
53:55That was a case report.
53:58And you can learn an awful lot
53:59from some case reports.
54:00We talk about the end of one
54:02trial in in oncology quite often,
54:04so there are things that we can do
54:07with case reports with small studies.
54:09Even better if we can band together
54:12internationally and actually get bigger
54:15studies put together on samples that
54:18we collect prospectively with a protocol.
54:21And then we've got a chance of getting
54:24some proper funding into it as well.
54:26And and we probably find we can
54:29go beyond pathology and pathology
54:31is simply the study of disease.
54:33So how we do it?
54:35Doesn't really matter whether
54:37it's with a microscope.
54:39Or an MRI.
54:41Or sequencing device doesn't actually matter.
54:46What we what we need to do is
54:48know what our questions are
54:50and what we were trying to do.
54:52Classification in its own right
54:54taxonomy is actually something that
54:56I think matters and which we can.
54:59We can almost certainly do better
55:02with than we are at the moment
55:05and get people to fund.
55:07So I think the future is
55:10pretty bright actually.
55:11That said,
55:11you can still do an awful lot of
55:13the microscope there, graticule.
55:16Uhm? Thank you, there is a
55:19question posted in the chat.
55:22Is there an electronic
55:24version of these books?
55:25For example, is the database
55:28behind it is available for download
55:31and use in electronic systems?
55:34Is there an API for querying the
55:38latest version of the database via
55:41a public or paid Internet access?
55:45Right, so that that that's a
55:48that's a multiple question.
55:49It's a good one too.
55:51The answer is that it's not easily available,
55:54it is a SQL database.
55:56So theoretically, yes, you can do it.
56:00We've got the, uh, electronic system
56:03going in terms of the website,
56:06so the website is is searchable,
56:08so you can do searches on
56:10the website and it's not.
56:12A sophisticated, not sophisticated,
56:13perhaps as it might be.
56:15But it could become more sophisticated,
56:18and I think we'd be very open
56:20to working with people on that.
56:23In terms of what we have to safeguard,
56:25clearly we have to safeguard the
56:26income of the of the program.
56:28'cause if we don't,
56:29we've got a problem.
56:31But there are.
56:34You know there are examples of where
56:36that's been done very successfully,
56:37for instance Genomics England have done
56:39it with their data so you can access that.
56:42It's just been done as well for a
56:45very large pathology set up in the
56:47middle of England called Path Lake,
56:49which has half a million annotated
56:52whole slide images. So there are.
56:57There are ways of doing it.
56:58The short, the short question.
57:00The short answer though.
57:01Sure.
57:03So I say no more questions and
57:06it's time to wrap up and come.
57:09I would just remind the audience that
57:13you are welcome to volunteer contact Dr,
57:17Cree and volunteer.
57:20In data analysis, or in any way you
57:24think your talents can be used.
57:28I hand over to Doctor Lu our chair.
57:32Yeah, thank you and I will
57:34share my screen now.
57:36Thanks talk to create for your
57:38coming and I I think unfortunately
57:40we couldn't do it in person and
57:43couldn't present you some token
57:45to you know on behalf of the
57:47department and our faculty.
57:48So I think you know we come up with this.
57:53Play, you know
57:55it will mail it will. So this is,
57:57uh, we're not just giving you a
57:59virtual plug, it's real well alright
58:01yeah it will be a real life this is a
58:03picture of that and I hope you know this
58:05we just you know really you
58:08know express our gratitude
58:09and if we are leadership for this,
58:11you know you know great book
58:13which is making such immense
58:15impact on pathology practice
58:18and I think it also this is
58:20just a reminder, you know
58:21we are ready to, you know to work with
58:23you to collaborate. And best wishes for
58:26the next. You know, the 5th
58:27edition of the Blue Book and
58:30thank you very much and saying,
58:31you know, hopefully we can
58:32see you in person sometime.
58:34You know either in your place
58:36or in our place in the future.
58:38Thank you, appreciate and I will say it.
58:40You thanks doctor.
58:41You know Manju Prasad,
58:43you know she this is her
58:45idea. So I'm just kind of presented
58:47her idea in this is great that
58:49you know you know idea to get. You
58:51know they will this you know present.
58:55Thank you very much indeed.
58:56It's been a great honor
58:57and a privilege to be here,
58:58and I've enjoyed talking to
59:00some of you over the afternoon
59:02and my afternoon anyway.
59:04And I hope the rest of the day for you
59:07goes well and I'm sure we will be in touch.
59:10Thank you.
59:11OK, thank you very much.
59:12Thank you.