Role of Pathology in Cancer

October 11, 2021

Information

October 10, 2021

Yale Cancer Center

visit: http://www.yalecancercenter.org

email: canceranswers@yale.edu

call: 203-785-4095

ID6988

To CiteDCA Citation Guide

00:00Funding for Yale Cancer Answers
00:03is provided by Smilow Cancer
00:05Hospital and AstraZeneca.
00:07Welcome to Yale Cancer Answers with
00:09your host doctor Anees Chagpar.
00:12Yale Cancer Answers features the
00:14latest information on cancer care by
00:16welcoming oncologists and specialists
00:17who are on the forefront of the
00:20battle to fight cancer. This week,
00:22it's a conversation about the
00:23role of pathology and cancer
00:25with Doctor Angelique Levi.
00:27Dr. Levi is an associate professor of
00:29pathology at the Yale School of Medicine,
00:31where Doctor Chagpar is a
00:33professor of surgical oncology.
00:36Angelique, maybe we can start off
00:38by you telling us a little bit about
00:40yourself and what it is you do.
00:42So as a pathologist I am anatomic
00:45and clinical pathology trained
00:47so a PCP for short and
00:50have received fellowship training
00:52and in Cytopathology,
00:55a subspecialty of the discipline.
00:58I have also received some extra
01:01training and expertise in GU pathology,
01:04all of which I did
01:06way back at Hopkins and
01:10it's a combined program.
01:13If you do both anatomic
01:15and clinical pathology
01:16it used to be five years and now is
01:184 but you can do one of the two
01:20disciplines for a little less time now.
01:23And the anatomic focuses
01:27mostly on the study of tissue
01:30and working with a microscope,
01:31fluids and cells,
01:33whereas the clinical pathology focuses
01:36a bit more on laboratory testing,
01:39blood tests for example.
01:41So let's dive a little bit more into that.
01:44I mean, when we think about the role
01:47of pathology and cancer automatically,
01:49our brain kind of goes to, Oh yeah,
01:52it's the pathologists who kind
01:53of look at the biopsy and tell
01:55me whether or not I have cancer.
01:57Can you flesh out a little bit
01:59more about what it is you do and
02:01how you come up with that answer?
02:03I mean everything hinges on what you say,
02:07how much pressure is that,
02:09and how do you actually come
02:10up with the correct diagnosis?
02:13It's certainly a team
02:15from the very beginning,
02:17patients will go to either a hospital
02:19or a physician office and will have
02:23a procedure done so the procedure
02:25could be either a Pap test,
02:29screening test for cervical cancer,
02:31it could be a fine needle
02:33aspiration of a breast mass,
02:34or it could be a surgical procedure
02:37in the operating room where
02:39a tumor or an organ is removed, so
02:43all of those tissues come to the lab from
02:47those scenarios and in the lab, the
02:51histology component is where that
02:55tissue is transformed into a medium
02:58where it is put onto a glass slide and
03:03that process itself is quite intense.
03:06We have pathology assistants who help
03:09in the gross examination of these
03:11tissues when they come to the lab,
03:14especially the larger ones where they
03:17may note sizes of lesions they may sample.
03:21Areas that are critical,
03:23close to margins, etc.
03:25And those sections are then
03:27submitted in cassettes and processed,
03:30in an automated lab in a way that
03:33they are sliced and stained and put
03:37on glass slides for pathologists
03:39to then review at the time of a
03:43case review and in community practice,
03:45often it is just a pathologist,
03:47but here at academic centers we
03:51have trainees, residents,
03:52who are involved in that process.
03:54We have many sets of eyes that
03:58we call preview slides and then
04:00the pathologist sits down at a
04:02microscope to sign out.
04:03And that's actually transforming as well.
04:05Soon we might say we don't sit
04:07down at a microscope to sign out,
04:09but we may sit at a computer screen
04:13if we transform into the digital era,
04:16but we're not quite there yet.
04:22Then with a microscope is where we
04:26really do what we were trained to do,
04:28and you use your trained eye to look
04:32at the morphology of the tissue and
04:35see where it differs from what you
04:38have trained yourself to know what's normal.
04:41So identifying what's abnormal
04:43disease and in that then deciding
04:47whether it's cancer or not.
04:49So not every disease is cancer,
04:51and it's important in some cases
04:53where the presumption clinically
04:54might be a mass because of cancer,
04:56it's a really important piece to
04:59be able to say this isn't cancer,
05:01and so therefore no treatment is necessary.
05:04But at a Cancer Center,
05:06many of the referrals that come here often
05:10perhaps already with a preliminary
05:12diagnosis on a small biopsy of
05:15cancer and then our job sometimes,
05:17as pathologists, in a larger procedure, or a
05:20resection is to then go ahead and stage that,
05:24which means assign some more
05:27parameters around that diagnosis.
05:29So not only is it cancer,
05:31but it's a type of cancer that
05:34you want to kind of classify.
05:37It's given a grade as we call it,
05:40well differentiated, poorly differentiated.
05:43It might be given certain
05:46other parameters regarding size or margin.
05:49Different cancers have different
05:51parameters that are important,
05:52and all of those details are important
05:56in prognosis prediction and then
05:58treatment and usually
06:00associated then with outcome.
06:03So I want to pick up on a few
06:05things that you said there.
06:06So one was this whole process
06:09that really goes on that
06:12many people who have never stepped into a
06:15pathology lab might not know about which is
06:17when you have a biopsy done
06:20and your surgeon, your radiologist,
06:23whoever has done the biopsy,
06:25sends that specimen away.
06:28Oftentimes, it's the greatest
06:29amount of patient anxiety waiting
06:31for that result to come back.
06:34And sometimes it can take a few days,
06:36but there is all of this
06:39preprocessing that needs to go on.
06:41Can you give us a sense of how long
06:45these biopsy results can sometimes take,
06:48and why it's important to really be
06:51patient and wait for your pathologist
06:52to give you the right answer because
06:55as you say so much of treatment really
06:58rests on what the pathologist says.
07:01Absolutely, that pre-analytical phase
07:02that you're talking about is a big part
07:05of our processing in the lab and
07:08that's kind of a traditional laboratory
07:10setting where you know pathologists when
07:13we talk about where do you work,
07:15you work in the lab, well no,
07:16we actually work mostly in our offices,
07:18but much of what's happening
07:19before we even see that glass slide.
07:29An Accessioner is the first person in the
07:32laboratory that basically does the
07:34patient registration that assigns that
07:38specimen a unique number.
07:40Every specimen in pathology is assigned a
07:42unique number and that's how we identify it.
07:45The patient information,
07:46clinical identifiers are then entered,
07:49and that's a really important step in
07:52terms of specimens being identified
07:54properly and assigned to the right person.
07:57That is the first thing that happens and
08:00the next step is it goes to the gross
08:05Histology bench and so for small biopsies
08:09that are cores or maybe liquid,
08:11or a pap smear,
08:13just single cells,
08:14fixation is something that
08:16doesn't take as long,
08:18so fixation is something that
08:20happens in different chemicals,
08:23alcohol and or formalin.
08:26Now when these tissues are larger,
08:29as in the case of a large tumor
08:31or resection or a large organ,
08:33that fixation process can
08:37happen over a 12 hour period.
08:39Sometimes overnight,
08:40so for example,
08:41a prostate that is removed whole
08:44or a large breast excision,
08:48those are examples of tissues that
08:49take a long time to fix in formalin.
08:52So before those sections can even be
08:54taken to embed in those paraffin blocks,
08:57that process has to happen,
08:59and it's critically important for
09:01that process to happen
09:03because these tissues need to be
09:06able to be examined in sections
09:08in a way where the margins
09:10and all of those distinctions
09:12between things that are critically
09:14important for patient care,
09:16whether the person gets radiation
09:18or not is the margin positive.
09:21Those delineations are critically
09:23dependent on that fixation step,
09:25and that step is where we really
09:28need to wait, and we can't rush it.
09:30So we have some technologies,
09:33microwave assistance and other things,
09:35but in that process there
09:39are still very manual
09:41pieces that take time and then
09:43by the time that slide comes out,
09:47if the surgery was on a Monday,
09:49that glass slide may not even
09:51come to a pathologist's desk until
09:54the following afternoon.
09:56And if that following afternoon
09:58is the first time a pathologist
10:00is looking at a cancer,
10:02whether it's a complicated case or even
10:06a standard morphologic diagnosis of,
10:09let's say, breast cancer, there are still
10:11additional tests that will have to get done,
10:14and so those tests will include
10:16immunostains and other markers
10:18that are all very important that
10:20need to be included in the report.
10:25So a lot of those those markers are
10:28things that we have to then order,
10:30and again it's another day
10:33or overnight processing,
10:34and so each of these steps
10:37requires kind of another decision
10:39and potentially another test
10:42or stain or molecular marker, for example.
10:45So important for people not
10:48to rush the pathologist because as
10:50I tell my patients,
10:53everything rests on what they say.
10:56But having said that,
10:58many people nowadays are
11:01talking about second opinions,
11:03either a second opinion
11:04from their clinician,
11:05but also getting their pathology that
11:08may have been reviewed at one institution
11:10re-reviewed at another institution.
11:12So for example,
11:13if they get a second opinion,
11:15their outside pathology is often re reviewed.
11:19So can you talk about the importance of
11:21that and how often do pathologists disagree?
11:24I mean, are these diagnoses
11:26things that are black and white?
11:28That is pretty crystal clear when
11:30you see a cancer that it's a cancer.
11:32Or are there some nuances that
11:34allow for some variability in
11:36terms of pathologic opinion?
11:38I'll start by saying second opinions within
11:42any scenario are always a good thing.
11:46I think for another set of eyes to
11:48take a look at a cancer case is
11:54always good and in the vast majority of
11:57cases a confirmation is what you'll find.
12:00The confirmation of the original diagnosis.
12:04It becomes more important in certain
12:07scenarios, so certain cancers
12:10have required subspecialty training
12:13that not all pathologists have,
12:16where you practice,
12:18where you've trained,
12:20and what you've become an expert
12:22in really does matter and standards
12:24are different for different places.
12:27In the community setting,
12:28while there's very high
12:30standards of care there,
12:32they may not always see all of the unique
12:36rare tumors that we might have in a tertiary
12:40academic center.
12:42Whereas in an academic center like Yale
12:45we would be able to kind of explain a bit
12:48more if there are nuances to a tumor.
12:51So black and white,
12:53yes cancer or not,
12:54in the vast majority of cases.
12:56But for challenging cases,
12:58I think second opinions are definitely
13:02helpful with expert review and consensus.
13:05Daily Conference is something
13:07that is part of our routine,
13:10not always in all practices.
13:12So it's important to kind of
13:14understand the nuances of pathology.
13:17We're going to pick up this
13:19conversation right after we take a
13:20short break for a medical minute.
13:22Please stay tuned to learn more about
13:24the role of pathology in cancer
13:26with my guest Dr. Angelique Levi.
13:28Support for Yale Cancer Answers
13:30comes from Smilow Cancer Hospital,
13:32where a dedicated team approach
13:34is used to diagnose liver cancer
13:36early when treatment is optimal and
13:38new, more effective treatments are
13:40being developed. To learn more visit
13:45Yalecancercenter.org/liver
13:48Over 230,000 Americans will be diagnosed with lung
13:51cancer this year and in Connecticut
13:53alone there will be over 2700 new cases.
13:56More than 85% of lung cancer
13:59diagnosis are related to smoking and
14:01quitting even after decades of use
14:04can significantly reduce your risk
14:06of developing lung cancer. Each day,
14:08patients with lung cancer are surviving
14:10thanks to increased access to advanced
14:13therapies and specialized care,
14:15new treatment options and surgical
14:16techniques are giving lung cancer
14:18survivors more hope than they
14:20have ever had before.
14:21Clinical trials are currently
14:23underway at federally designated
14:25Comprehensive cancer centers,
14:26such as the battle two trial at
14:29Yale Cancer Center and Smilow
14:30Cancer Hospital to learn if a drug
14:32or combination of drugs based on
14:35personal biomarkers can help to
14:36control non small cell lung cancer.
14:39More information is available at
14:42yalecancercenter.org. You're listening
14:44to Connecticut Public Radio.
14:47Welcome back to Yale Cancer Answers.
14:49This is doctor Anees Chagpar
14:50and I'm joined tonight by my
14:52guest doctor Angelique Levi.
14:54We're talking about the role of
14:56pathology in cancer and Angelique,
14:58you know one of the things that you
15:00mentioned before the break that I
15:02was interested in is that you said
15:03we are getting close to a
15:06time when pathologists might not be
15:09looking down the microscope anymore.
15:11They might be looking at a computer
15:13screen and that brought to mind
15:16this whole concept of digital
15:18pathology and potentially the role
15:20of artificial intelligence in helping
15:23pathologists make that diagnosis.
15:25You talked a little bit before the
15:27break about some of the nuances.
15:29Can you talk a little bit about
15:32where you see digital pathology and
15:35the role of AI kind of playing
15:37into pathology as we move forward?
15:39Absolutely, the landscape is already
15:44changing and the field is rapidly evolving.
15:50And pathologists, I think are
15:54definitely stepping up and wanting to
15:58not just join this era of digital and
16:03artificial intelligence as you say,
16:05machine learning, but hopefully
16:08also take a role in leading that
16:11charge and for pathology there are
16:14so many potential applications as
16:16with everything AI is everywhere.
16:19We don't necessarily appreciate
16:21it from our phones and our apps,
16:23or for many the interfaces we do each day,
16:27but it's a tool no different
16:30than for pathology,
16:32maybe an immunostain and
16:34molecular marker or genetic
16:35profile and how we use that tool
16:39is largely dependent on
16:42what help or guidance a particular
16:45practice might be looking for.
16:48One example of AI and pathology as
16:51you mentioned or alluded to would be,
16:53helping to make a diagnosis
16:56or grading a tumor.
16:59So an area of study that I
17:03have pursued in GU pathology
17:06and in prostate cancer
17:07this is a common application now
17:11and there are already software
17:15companies that are promoting
17:19AI programs and software that can
17:23reliably help predict grades or
17:26Gleason scoring of prostate cancer.
17:29But it's not that simple.
17:34Depending on the cancers that might be
17:38seen in a given institution,
17:40whether it's more common,
17:43lower grade, or in a tertiary care center,
17:46much more complicated,
17:48higher grade,
17:49algorithms are kind of taught to
17:52answer a specific question or grade.
17:55So if you're looking for well differentiated
17:58prostate cancer 3 + 3 Gleason score,
18:01that might be one training set,
18:04whereas if you're looking for
18:07high grade prostate cancer,
18:09that is
18:11amenable not to resection,
18:13but further treatment,
18:14that might be another training
18:17software kind of algorithm,
18:18so much depends on
18:21the question being asked,
18:22and it's not just help in grading,
18:24but it could also just be help in detection,
18:28so different programs can be
18:31taught how to do different tasks,
18:34and another program might be in
18:37a better setting for community
18:39practice not to miss cancer
18:42as much as
18:44focusing on the grade because
18:47detection and preventing false
18:49negatives would really be the key
18:53perhaps in a Community setting
18:54with a lower cancer rate,
18:56whereas at the tertiary setting something
18:59that would be more helpful is perhaps
19:03an AI software algorithm that not
19:06just helps with detection or grade,
19:09but maybe with prognosis.
19:10And that's really the key,
19:13trying to discern what this AI can help
19:16with and how we'd like to apply it,
19:21tailoring the solution to the problem.
19:22But one of the questions is this.
19:26Are these technologies in use now?
19:28And is there a way for patients to
19:31know whether a particular pathology
19:33department is using that or not?
19:35For example,
19:36if I just had a biopsy at my
19:39Community Hospital and I want to make
19:42sure that they didn't miss a cancer,
19:44should I expect that they have that kind of
19:48technology that can help the pathologists?
19:51And if I'm not sure,
19:52is there a way to find out?
19:54There's always a way to find
19:56out and certainly just calling that
19:58pathologist on the bottom of the
20:00report would be the first step.
20:05Or wherever those procedures are done would
20:07certainly know within the department,
20:09I would say we're still on the cusp.
20:18I think right now in tertiary care centers,
20:22there are many kinds of testing and
20:25research scenarios and these
20:27are all kind of sprouting
20:30up now and it's not to
20:33be expected I would say because
20:36it requires so much investment and
20:39infrastructure.
20:47Whether it's cloud based
20:51memory or machine or human time,
20:55we can't expect that to all be there.
20:58Now I would say you know,
21:01in the future 5 to 10 years,
21:0410 to 15 years, I think
21:05then we can start
21:07to see where these
21:10applications are best suited,
21:12and imagine with all of this investment
21:17it would probably be helpful as a QC measure.
21:21You know there are always reimbursement
21:24codes for things that are additive,
21:27whether it's a stain or
21:29whether it's AI assisted.
21:30So I imagine in the future it would also be
21:33part of a report and so you know we're not
21:37there yet, but it does take a lot of time,
21:40infrastructure,
21:40and money frankly,
21:42and so until those costs come
21:45down or those partnerships are
21:47established, things
21:49may be commercially available
21:53at a price that is affordable for you.
22:04The other thing that is here now
22:08more and more in the cancer world is
22:10this whole concept of personalized
22:12medicine and so many clinicians are
22:15really now trying to unlock and
22:18understand the genomics of cancers.
22:20And we've certainly had guests on this
22:22show who talk about doing stains that
22:25look at a number of different
22:28genetic and genomic mutations that
22:31actually help in figuring out how
22:34a particular tumor may be treated.
22:37Is that done at the local pathology lab?
22:40What's the role of the pathologist in that?
22:43How do you decide which of these
22:46markers really needs to be done?
22:48What's the cost and is that standard
22:51of care or is that something that
22:54patients need to really individualize?
22:57So at the local level I don't
23:00think it's necessarily standard of care.
23:03Certainly immunostains,
23:09certain markers that are common
23:12to lay folks would be,
23:15we talk about estrogen and progesterone
23:17receptors for breast cancer.
23:18For example, ER,
23:20PR and certain molecular markers.
23:22I think in Community practice
23:25the idea is
23:27to partner often with
23:29another lab.
23:31Whether it's a tertiary center,
23:33a commercial lab that offers those
23:36tests because they are not able to
23:39have access to all of that in house,
23:42and so a lab like
23:45ours comes into play,
23:47where if we have something to offer,
23:50we can partner with other network hospitals,
23:54community hospitals,
23:55even other labs that might not have
23:58the same volume we do in a Cancer
24:00Center to provide all of these highly
24:03specialized tests that without a certain
24:05volume it's not affordable to run.
24:08So I think the same thing holds for
24:12additional molecular assays.
24:15Panel genetic profiling those are highly
24:20specialized areas and fields that
24:23without partnering with another kind
24:26of tertiary care center or larger
24:29lab specifically geared towards that,
24:32I think it's not
24:34expected at the local level.
24:36So are the decisions about what
24:39additional tests need to be done,
24:41so additional molecular
24:42tests and so on, EGFR VEGF,
24:45various mutational panels
24:46and so on are those decisions
24:50made by the pathologist, by the
24:53treating clinician, by a group?
24:56How are those decided?
24:58I think in the Community level
25:01the oncologist drives a lot of that because
25:04the oncologist sees on that leading edge
25:06what the potential drugs that are
25:10available that are targeted to a
25:14particular molecular change, and so in
25:17the Community setting,
25:19I think the oncologist takes that
25:22role more so in asking a pathologist,
25:25hey, there's a new drug and it targets
25:29this molecular marker.
25:30Is that something you do in your lab?
25:34Or is it something we can send out for?
25:35And then you know,
25:37the pathologist facilitates that.
25:38And so that I think happens
25:40more on the Community side,
25:42whereas I think in the tertiary care setting,
25:45like here,
25:45I think it is a bit more of a
25:48collaborative effort because there
25:50are there are the pathologists
25:54here who are doing those genetic tests
25:56and so we also have our tumor boards
25:59that while they have been outside
26:01at the Community level as well,
26:03I think in a Cancer Center,
26:05the tumor boards really are
26:07putting everyone at the table.
26:09Who has that subspecialty expertise?
26:12And so I think it's a bit more
26:13of a collaborative effort.
26:15And if there's something that is
26:18clinically warranted or a new drug,
26:21I think the pathologists here in a
26:25tertiary center are able to create
26:28these answers to some of
26:30those questions or research them,
26:32or they're already a line of research here
26:35in the department or collectively.
26:41Which segues nicely into you,
26:44know, one of my last questions,
26:45which is what are the exciting areas of
26:48research in pathology and cancer?
26:50I mean, it seems like pathology
26:53is so central to what we do.
26:56Are there some exciting developments
26:58that you see coming down the Pike
27:00in terms of pathology and cancer?
27:03Well, I definitely think the
27:06digital pathology component and
27:09the artificial intelligence piece
27:11is very exciting.
27:14It's entirely a new platform
27:16and revolution, so to speak.
27:18It's something that can be applied
27:20to all of the tools that we have
27:23and then it's a tool on its own.
27:25So what I mean by that is the
27:29ability to work with digital images,
27:33whether it's radiology or
27:36scanned pathology slide and
27:39with that scan slide use metrics or
27:44segmentation to detect changes
27:47that maybe even the human eye can't.
27:50And maybe it's not just about morphology,
27:53it's just a whole other level
27:56of detection
27:58in addition to our molecular
28:01assays and genetic profiles,
28:03is something that can on its own be
28:06additive and the exciting pieces when it
28:09is also its own prognostic indicator,
28:12and so we're always interested in
28:15knowing more about the meaning of
28:17the cancer and what effect
28:19that has on outcome and prognosis,
28:22and AI really has the potential
28:24to help each of these special
28:28techniques that we use and the
28:30ability to stand on its own.
28:32Doctor Angelique Levi is an
28:34associate professor of pathology
28:36at the Yale School of Medicine.
28:38If you have questions,
28:39the address is cancer
28:40Answers at yale.edu and past editions of
28:43the program are available in audio and
28:47written form at yalecancercenter.org.
28:48We hope you'll join us next week to
28:51learn more about the fight against
28:53cancer here on Connecticut Public Radio.
28:54Funding for Yale Cancer
28:56Answers is provided by Smilow
28:57Cancer Hospital and AstraZeneca.