Nomograms and Gene Expression Profiles for Cutaneous Malignancies

Transcript

• 00:00Program tonight brought to you
• 00:02by a multidisciplinary
• 00:03group to hear from the
• 00:04Smilow Cancer Center encompassing expertise
• 00:07in all of cutaneous
• 00:09oncology.
• 00:11We're gonna start off the
• 00:12session, which really
• 00:14focuses
• 00:15on gene expression and cutaneous
• 00:18malignancies as we've noticed a
• 00:20real uptake in the use
• 00:21of this. And there's been
• 00:22some really great recent data,
• 00:25a flood of it actually
• 00:26recently from randomized controlled trials,
• 00:28as well as some opening
• 00:29of some clinical trials around
• 00:31the country.
• 00:32So we're going to begin
• 00:33our discussion today,
• 00:35with doctor Swoozie.
• 00:37If you'd like to get
• 00:37us started, please.
• 00:41Sure.
• 00:42Just give me one sec.
• 00:46So hi, everyone.
• 00:48Oh, did it stop? Hold
• 00:50on. My screen share stopped.
• 00:52Let me try again here.
• 00:58Here we go.
• 01:00Hi. I'm Katie Swozzi. I'm
• 01:02a Mohsurgent here at Yale,
• 01:04and I'm gonna be starting
• 01:05us off tonight by discussing
• 01:06the role of nomograms and
• 01:08introducing,
• 01:10the idea about gene expression
• 01:11profiling in cutaneous squamous cell
• 01:13carcinoma
• 01:14with a focus on how
• 01:16these tools can help refine
• 01:17risk assessment and clinical decision
• 01:19making.
• 01:20I'll review briefly traditional staging,
• 01:23introduce nomogram based prediction,
• 01:26particularly the risk calculator, and
• 01:28discuss how molecular tools like
• 01:29GEP may further refine risk.
• 01:32So although most cutaneous squamous
• 01:34cell carcinoma have excellent prognosis,
• 01:36a subset behaves aggressively and
• 01:38the consequences of missing high
• 01:40risk disease are significant.
• 01:42Rates of unfavorable outcomes are
• 01:43about three to five percent
• 01:44for local recurrence, two to
• 01:46five percent for nodal mets,
• 01:48and one point five to
• 01:49three point five percent for
• 01:50disease specific death. And despite
• 01:52these relatively small percentages, the
• 01:54absolute
• 01:55mortality from cutaneous squamous cell
• 01:56carcinoma is substantial,
• 01:59meeting at least meeting,
• 02:00but, and by most models
• 02:02exceeding that of melanoma.
• 02:04So I did talk about
• 02:06staging in the last one
• 02:07of these sessions, but I
• 02:08did wanna briefly bring it
• 02:09up here. So first here,
• 02:10I'm showing AJCC eight.
• 02:13Staging and, the eighth edition
• 02:16for head and neck cutaneous
• 02:18squamous cell carcinoma
• 02:19attempted to improve risk stratification
• 02:21by expanding the t three
• 02:23group to include deep invasion,
• 02:27perineural invasion, and minor bone
• 02:29erosion.
• 02:30And one of the major
• 02:31improvements compared with AJCC seven
• 02:34was that core outcomes are
• 02:35now concentrated in t three
• 02:36and t four tumors
• 02:38rather than across a heterogeneic,
• 02:40T2 group.
• 02:42However, even with AJCC8,
• 02:44there remains substantial
• 02:45variability in outcomes.
• 02:47So the Brigham and Women's
• 02:48Staging System was developed to
• 02:50address this heterogeneity.
• 02:52This model,
• 02:54identifies four major risk factors,
• 02:56diameter,
• 02:56depth, beyond fat, perineural invasion
• 02:59of greater than point one
• 03:00millimeters, and poorly differentiated histology.
• 03:03And tumors are then classified
• 03:05by their number of risk
• 03:06factors into t one, t
• 03:08two a, t two b,
• 03:09and t three.
• 03:10And Brigham and Women's Staging
• 03:11shows better risk stratification
• 03:14at, t two, t three,
• 03:15and here you can see
• 03:16that t two b tumors
• 03:18have about a twenty percent
• 03:19risk of local recurrence and
• 03:21nodal metastasis
• 03:22with t two a,
• 03:23ten percent or below.
• 03:25And you can see there's,
• 03:27less stratification,
• 03:28in AGCC
• 03:29eight, between t three,
• 03:32and t two.
• 03:33And so this improved risk
• 03:35stratification, but staging remains categorical
• 03:38rather than individualized.
• 03:40So NCCN
• 03:41guidelines build further on staging
• 03:43incorporating additional
• 03:44high risk features and patient
• 03:46specific features such as, you
• 03:48know, is the tumor primary
• 03:49recurrent? Is the patient immunosuppressed?
• 03:52Is the location on previously
• 03:54irradiated skin? What is the
• 03:56growth rate of the tumor?
• 03:58And also incorporates different,
• 04:00histologic subtypes.
• 04:03And NCCN classification is useful,
• 04:05but it still groups patients
• 04:06into low, high risk, and
• 04:08very high risk rather than
• 04:09generating individual probabilities.
• 04:13So nomograms,
• 04:15this is where nomograms are
• 04:16particularly helpful. So a nomogram
• 04:18is a statistical model that
• 04:20integrates multiple variables to generate
• 04:22an individualized
• 04:23probability of prosthesis
• 04:25or death.
• 04:26And unlike staging systems, nomograms
• 04:28produce continuous risk estimates rather
• 04:31than categories.
• 04:32And this is especially useful
• 04:33in gray zones. For example,
• 04:35tumors that fall between Brigham
• 04:37and Women staging t two
• 04:38a and t two b,
• 04:39which we commonly refer to
• 04:40as t two a plus.
• 04:44Nomograms can also be, presented
• 04:46graphically
• 04:47or embedded in calculators,
• 04:49making them very clinically
• 04:51practical. So,
• 04:52I briefly mentioned this nomogram,
• 04:55for cutaneous SCC that predicts
• 04:57the risk of recurrence after
• 04:58MOSE, which was published in
• 05:00twenty twenty two, combining clinical
• 05:02factors with two biomarkers,
• 05:04and here showing the type
• 05:05of graphical representation of the
• 05:07nomogram. But the one I
• 05:08wanted to focus on is
• 05:09the, risk calculator, which was
• 05:11developed by,
• 05:13number of doctor Krishnan and
• 05:14I's colleagues in the Mose
• 05:15College. And,
• 05:17it's a clinically practical nomogram
• 05:20used to predict metastatic risk,
• 05:22and it's also available in
• 05:24an app form that's very,
• 05:26user friendly.
• 05:27So I wanted to show
• 05:28you the risk calculator using
• 05:30a case I recently saw
• 05:31in clinic. So here's a
• 05:32patient who presented for a
• 05:33treatment of a cutaneous squamous
• 05:35cell carcinoma,
• 05:36the right preauricular region.
• 05:38And here was her PRIMOs,
• 05:41lesion and here's her post
• 05:42MOS defect. And so when
• 05:44you look at staging for
• 05:45her, her PRIMO size was
• 05:47one point five centimeters. She
• 05:49had very widespread perineural invasion,
• 05:51but it was all small
• 05:52caliber,
• 05:52and she did have poorly
• 05:53differentiated histology. So, by staging,
• 05:56she's either AJCCA
• 05:58T1 or T2A,
• 05:59but you just know this
• 06:00patient is not going to
• 06:02do well. And,
• 06:03that that staging is underrepresenting
• 06:05the biology of her disease.
• 06:06So let's put her in
• 06:08the risk calculator. So here,
• 06:09if we add,
• 06:11her criteria,
• 06:13here, you see she's gonna
• 06:14get credit for a perineural
• 06:15invasion for the depth of
• 06:17invasion.
• 06:18And,
• 06:19we look at her risk
• 06:20calculator, it's gonna reflect a
• 06:22lot higher than her staging,
• 06:25her risk of neurotoxicity
• 06:27being, over thirteen percent and,
• 06:30recurrence after mood was about
• 06:32twenty percent. So this patient
• 06:33will go to our multidisciplinary
• 06:35tumor board for consideration of
• 06:37adjuvant,
• 06:38therapy where her staging may
• 06:39not have triggered that alone.
• 06:42So I'm briefly just gonna
• 06:44talk about,
• 06:44GP because doctor Roach is
• 06:46gonna build on this. But,
• 06:48for cutaneous SCC, we have
• 06:49a forty gene expression,
• 06:51profile test that's clinically available.
• 06:54This was generated from primary
• 06:56cutaneous bimicile carcinomas with known
• 06:58outcomes, so retrospectives,
• 07:01we developed,
• 07:03and has not been prospectively,
• 07:06validated yet, but identifies three
• 07:08classes, class one, two a,
• 07:10and two b, the different
• 07:12likelihoods of developing metastasis,
• 07:14at three years.
• 07:15Oh, sorry, this slide looks
• 07:16weird.
• 07:18So, the three classes, Class
• 07:19I is the low biological
• 07:21risk and is about half
• 07:22the study population,
• 07:23Class two a is moderate
• 07:25biologic risk, and, when you
• 07:26look at the hazard ratio
• 07:28here, it's similar to the
• 07:29strongest traditional clinical pathologic factors
• 07:32like perineural invasion, for example.
• 07:34And class two b is
• 07:35very high risk with a
• 07:36greater than fifty percent risk
• 07:38of metastasis, and you could
• 07:39see that here.
• 07:42So,
• 07:43about a third of the
• 07:44cases will end up, t
• 07:46two a. And so,
• 07:47well, doctor Roche is gonna
• 07:48talk more about what we
• 07:49do, with those tumors, and
• 07:51this is where, you know,
• 07:54correlation with staging is also,
• 07:57important. And so, you know,
• 07:59what are the takeaways here?
• 08:00Cutaneous squamous cell carcinoma is
• 08:02a heterogeneous
• 08:04group, and, accurate
• 08:06risk stratification is critical.
• 08:09Nomograms such as risk provide
• 08:11individual risk prediction and particularly
• 08:13helpful in these gray zone
• 08:14cases, and gene expression profiling
• 08:17adds biologic risk information that
• 08:19may further refine management. And
• 08:20to further speak about that,
• 08:20I'm gonna turn it over
• 08:21to
• 08:29Thank you, Katie.
• 08:31Alright. Let's pull mine out
• 08:32here.
• 08:34Share. Presenter mode. Okay.
• 08:37Alright. So I'm Ansley Roach.
• 08:39I'm a head and neck
• 08:40surgeon here.
• 08:42I'm gonna be just speaking
• 08:44specifically about,
• 08:46this forty GEP panel and,
• 08:49also as it pertains to
• 08:50the head and neck.
• 08:51But talking a little bit
• 08:52about its development, because I
• 08:53think it's important to just
• 08:54get an understanding of what
• 08:57of what this, how this
• 08:58was validated. So just to
• 09:00recap some of the things
• 09:01that Doctor. Swozi mentioned,
• 09:03the high risk features that
• 09:05we see in these patients
• 09:06can vary or can be,
• 09:08attributed to the tumor characteristics
• 09:10themselves. And I've listed them
• 09:12here. Essentially, anything greater than
• 09:13two centimeters depth of invasion
• 09:15beyond the subcutaneous fat or
• 09:17six millimeters.
• 09:18Higher risk areas, so specifically
• 09:20the head and neck, hands,
• 09:21feet, the pre tibial region,
• 09:23and the anal genital re
• 09:24reach, areas,
• 09:26poorly defined borders. Reshare your
• 09:28screen? I think it's not
• 09:29showing up. Sorry.
• 09:31That's okay. Thank you. Sorry.
• 09:32Let's see.
• 09:35Okay.
• 09:37Share. Sorry about that.
• 09:40What about this? Is that
• 09:41good? Okay. Is it in
• 09:42presenter mode? Or No. So
• 09:44if you go up and
• 09:45you could put it in
• 09:46presenter mode.
• 09:48Do I do that through
• 09:49right there.
• 09:52If you hit slideshow and
• 09:53then say,
• 09:57view
• 09:58I'm not saying Exit animation.
• 10:03What about this?
• 10:05No. It just doesn't know.
• 10:06Stopped. Okay. Sorry. So I'm
• 10:08just gonna do it old
• 10:09school.
• 10:11Yes. That? Okay. I'm not
• 10:13okay. Great. Thank you. Sorry
• 10:14about that.
• 10:16And then, also
• 10:18indications of,
• 10:20perineural invasion such as pain
• 10:22or paresthesias or, paresis or
• 10:25paralysis,
• 10:27histopathologic
• 10:28features, so just different histopathologic
• 10:30subtypes of squamous cell carcinoma,
• 10:32certainly the PNI and LVI
• 10:34that Doctor. Swozi mentioned, and
• 10:35then patient factors. These all
• 10:37place patients at high risk,
• 10:39for metastasis, local recurrence of
• 10:42squamous cell carcinoma.
• 10:45As in other disease states
• 10:46such as breast cancer,
• 10:49there have been,
• 10:51a a burgeoning
• 10:52field of gene expression profiling,
• 10:56in which a essentially, a
• 10:57proprietary algorithm is used,
• 11:00to predict metastatic risk or
• 11:02risk of recurrence.
• 11:04In this there's one that's
• 11:05commercially available for squamous cell
• 11:06carcinoma. It's a forty gene
• 11:08panel.
• 11:09They use two gene expression
• 11:11signatures. There are thirty four,
• 11:13discriminant genes. They're all listed
• 11:15here. It's kind of a
• 11:16a variety of different sorts
• 11:17of genes they're looking for,
• 11:19MMPs,
• 11:20metabolic related proteins, transcription factors,
• 11:23and then some control genes.
• 11:25And these this testing can
• 11:27be done on archived FFPE
• 11:28samples, so it does not
• 11:29have to be on fresh
• 11:31tumor.
• 11:33The this test was,
• 11:35developed and as doctor Swozi
• 11:37said, in a retrospective fashion
• 11:39where,
• 11:41long term outcomes were available
• 11:43and evaluated as part of
• 11:45the validation cohort.
• 11:47It was a multi institutional
• 11:49study across con across the
• 11:51country,
• 11:52pulling, three twenty one patients
• 11:54for the validation cohort. It
• 11:56was powered to detect a
• 11:57hazard ratio of three year
• 11:58metastasis
• 12:00of two point one.
• 12:01It, included patients that were
• 12:03at least had at least
• 12:04one high risk feature, so
• 12:05at least
• 12:06BWH
• 12:08one, and then
• 12:09a slightly higher risk of
• 12:11metastasis rate for these patients,
• 12:13or rate of metastasis, sixteen
• 12:14point two percent, versus the
• 12:16general population, which ranges anywhere
• 12:18from two to five percent.
• 12:21Similar to the,
• 12:22survival curves that doctor Swozzi
• 12:24shared, you can see here
• 12:26that,
• 12:27the
• 12:28three year metastasis free survival,
• 12:31is separated out, stratified by
• 12:33these classes with class one
• 12:35a, again, the low biological
• 12:37risk,
• 12:38with a three year metastasis
• 12:39free survival about ninety one
• 12:41percent. Two a would be
• 12:43the moderate,
• 12:44moderate risk. That's about eighty
• 12:46percent through your survival. And
• 12:472b, as we would expect
• 12:49with the highest risk, biological
• 12:51behavior,
• 12:52through your metastasis free survival
• 12:54about forty four percent.
• 12:58Hazard ratio,
• 12:59it was powered to demonstrate
• 13:01this and hazard ratio,
• 13:03for three year metastasis,
• 13:05for 2a was two point
• 13:07four four, and then it
• 13:08increases tenfold. So three year
• 13:11risk of metastasis for 2B
• 13:12tumors is tenfold,
• 13:14greater,
• 13:16than class I. And then
• 13:17the disease specific survival disease
• 13:19specific death also is significantly
• 13:21increased, for both,
• 13:242A and 2B tumors.
• 13:26I think it's important to
• 13:27just look at how this
• 13:28tool was developed.
• 13:30It was,
• 13:32it was,
• 13:33deep learning,
• 13:35on training data that was,
• 13:37one hundred and twenty two
• 13:39patients.
• 13:40I think it's important to
• 13:42note that as you can
• 13:43see here,
• 13:45the AJCC class t three
• 13:47and t four tumors are
• 13:48quite underrepresented,
• 13:49specifically t four. There's one
• 13:51patient with a t four
• 13:52tumor.
• 13:53And I think the the
• 13:54same can be said for
• 13:55BWH staging, so a pretty
• 13:57small sample size for t
• 13:59two b and t three
• 14:00tumors.
• 14:02When they, trained their,
• 14:04their model,
• 14:06and then compared it or
• 14:07then validated against
• 14:09additional cohort of three twenty
• 14:11one patients,
• 14:12their,
• 14:13numbers of T3 and T4
• 14:15tumors were slightly increased, but
• 14:16still quite small. So just
• 14:18seven seven patients with T4
• 14:19tumors by AJCC eight and
• 14:22seven with T3 tumors by
• 14:23BWH.
• 14:25The,
• 14:26at the end of their
• 14:27kind of validation study, they
• 14:28found that about half of
• 14:29their patients had this class
• 14:31one low biologic risk. Class
• 14:322a was just under half
• 14:34with a very small proportion
• 14:36being class 2b, that very
• 14:37high risk.
• 14:39I think it's also interesting
• 14:40to look at some of
• 14:41their,
• 14:42some of their regression data.
• 14:45Their data failed to show,
• 14:47an increased,
• 14:49or a hazard ratio for
• 14:50three year metastasis for T4
• 14:52tumors.
• 14:53You would think that by
• 14:54AJCC criteria T4 tumors, you
• 14:56would see an increased risk
• 14:58of three year metastasis. It
• 14:59just makes sense. These are
• 15:00advanced stage tumors, but their
• 15:02data is not showing this.
• 15:04And the same is true
• 15:05for the BWH
• 15:06staging. Their t three tumors,
• 15:07there's not an increase or
• 15:09there there's no increased kind
• 15:10of hazard of, three year
• 15:12metastasis.
• 15:13When they kind of change
• 15:15the data a little bit
• 15:16and combine some of these,
• 15:18the the staging, they combined
• 15:20t one and t two
• 15:21and then combined t three,
• 15:22t four, and then they
• 15:24did find a,
• 15:26an increased hazard of three
• 15:28year metastasis
• 15:29of
• 15:30two,
• 15:31with,
• 15:32class 2b having a near
• 15:34tenfold,
• 15:35increased risk of three year
• 15:36metastasis.
• 15:38But important to note that
• 15:39advanced stage tumors are underrepresented
• 15:41in this validation study. So
• 15:42what raises the question of,
• 15:43is this a useful tool
• 15:45for these advanced stage tumors,
• 15:48T3, T4 by AJCC criteria.
• 15:51Efforts to look at all
• 15:53the available literature,
• 15:54have been made of meta
• 15:55analysis and systematic review, looked
• 15:57at all available,
• 16:00research at the time. This
• 16:01was a twenty twenty three
• 16:02meta analysis, and they were
• 16:04after they, looked through all
• 16:06available studies, they, found three
• 16:08studies that reached that met
• 16:10inclusion criteria representing just over
• 16:12a thousand patients.
• 16:13They did find similar three
• 16:15year metastasis free survival across
• 16:17classes, where
• 16:19similar to the data I
• 16:20presented earlier and what Doctor.
• 16:21Swozzi presented,
• 16:23class I anywhere from ninety
• 16:25one percent to ninety three
• 16:26percent, three year metastasis free
• 16:28survival.
• 16:29Two a is slightly worse
• 16:30at seventy six to eighty
• 16:31percent. And two b,
• 16:33the worst of the of
• 16:34the three with forty three
• 16:35to forty seven percent
• 16:37three year metastasis free survival.
• 16:41Looking at positive predictive value
• 16:43of the forty GEP,
• 16:45is actually a fairly,
• 16:47substantial positive predictive value, about
• 16:49fifty five percent,
• 16:51for class 2b
• 16:53when compared to class one
• 16:54and class 2a.
• 16:56And when you compare that
• 16:57to the positive predictive value
• 16:59of AJCC
• 17:00and BWH,
• 17:02the forty gene panel is
• 17:04higher. So the positive predictive
• 17:06value for AJCC was just
• 17:08about a third, thirty three
• 17:09percent, and for the Brigham
• 17:10and Women's, thirty six percent.
• 17:12So
• 17:13the positive predictive value for
• 17:15the forty gene, expression profiling
• 17:17test is does outperform AJCC
• 17:20and BWH.
• 17:22To look specifically at head
• 17:24and neck,
• 17:25a subset of the original
• 17:26validation cohort was analyzed.
• 17:29Similar researchers,
• 17:30on this study as well,
• 17:31but looking specifically at the
• 17:33head and neck.
• 17:34Similar Kaplan Meier curves for
• 17:36these with class I three
• 17:38year metastasis free survival ninety
• 17:39two percent. You can see
• 17:40the numbers here, seventy six
• 17:42for class IIa, forty four
• 17:43for class two b. So
• 17:44very similar to,
• 17:46looking at,
• 17:47squamous cell carcinomas of the
• 17:49entire body, when you look
• 17:50at this for head and
• 17:51neck cancers.
• 17:54The sites represented here,
• 17:57where lip and scalp actually
• 17:59had the highest rates of
• 18:00metastasis.
• 18:01The most common locations
• 18:03were the, ears, cheeks, and
• 18:06scalp.
• 18:07But when they looked at
• 18:08metastasis rate by location on
• 18:10the head and neck, there
• 18:11they found no difference, in
• 18:13metastatic events and,
• 18:15location of the head and
• 18:16neck.
• 18:18And then here you can
• 18:19see here again depicted the
• 18:20this is it's very similar
• 18:22to what, what I presented
• 18:23with the validation study. This
• 18:25when you extrapolate out or,
• 18:27analyze the head and neck
• 18:28patients, the say it's a
• 18:29similar positive predictive value of
• 18:31about fifty five percent,
• 18:33higher than the positive predictive
• 18:35value. And that's for the
• 18:35for the for the gene,
• 18:37gene expression profiling test, fifty
• 18:39five percent positive predictive value
• 18:42compared to just thirty seven
• 18:43percent for AJCC
• 18:45eighth edition for advanced stage
• 18:46tumors and the BWH,
• 18:49advanced stage tumors. So, again,
• 18:50it outperforms those two for
• 18:52the positive predictive value.
• 18:55So like the AJCC eighth
• 18:57edition and Brigham women's
• 18:59staging system, the forty gene,
• 19:01expression profile testing does class
• 19:04or it classified lower rates
• 19:06of tumor as high risk
• 19:07compared to the NCCN.
• 19:09As doctor Suozzi showed,
• 19:10NCCN
• 19:12breaks patients down by low,
• 19:13high, and very high.
• 19:15And just by definition,
• 19:18head and neck is high
• 19:20risk.
• 19:20And so this test actually
• 19:23kind of downstages,
• 19:25downgrades,
• 19:26patients of the head and
• 19:28neck,
• 19:28whereas the NCCN,
• 19:30would not do that.
• 19:31And then when they compared
• 19:32these the head and neck
• 19:34sites,
• 19:36to,
• 19:37cancers of all sites of
• 19:38the body, they found that
• 19:39a higher percentage of cases
• 19:41were at moderate risk, so
• 19:42at 2a,
• 19:43and that a lower percent
• 19:45was at, the low risk
• 19:47of of class one,
• 19:49implying a sort of shift
• 19:50to higher risk for head
• 19:51and neck region, which we
• 19:53know that to be true.
• 19:55So I'm just gonna demonstrate
• 19:57kind of the impact that
• 19:58this could have in some
• 19:59clinical scenarios. So these are
• 20:01case studies that,
• 20:03two cases from a case
• 20:05series published in two thousand
• 20:06and two, two very similar
• 20:08patients from just kind of
• 20:10their history.
• 20:11So the first is a
• 20:12sixty five year old male
• 20:13with a double transplant,
• 20:15with, who developed this sort
• 20:17of small one point three
• 20:18centimeter papule on the temple.
• 20:20By AJCC, it's a stage
• 20:22one. BWH, it was a
• 20:23stage two a.
• 20:26He underwent Moe's, surgery,
• 20:28several stages.
• 20:30There was a an initial,
• 20:33initially, all margins were determined
• 20:34to be negative, but a
• 20:35subsequent analysis showed a small
• 20:37focus of,
• 20:39of cancer remaining in the
• 20:40tumor bed. Patient was notified,
• 20:43discussed additional treatment with radiation
• 20:45oncology. He declined.
• 20:47And,
• 20:48at the time of follow-up,
• 20:49he,
• 20:50was without disease.
• 20:52They tested his tumor, and
• 20:54it was a class one,
• 20:56for the forty GEP.
• 20:59A very similar patient, a
• 21:00sixty nine year old male,
• 21:01also liver transplant, several month
• 21:03history of a similar size
• 21:05lesion in a similar location.
• 21:07This one's exophytic,
• 21:09in its appearance, but similar
• 21:10staging, AJCC eighth edition one
• 21:12and then the BWH
• 21:14t two a.
• 21:15So this patient underwent,
• 21:17Mohs micrographic surgery, two stages,
• 21:20margins were clear,
• 21:22very short interval recurrence at
• 21:24about four months.
• 21:27He consulted with, radiation oncology.
• 21:30He received
• 21:31he received radiation to the
• 21:33temple.
• 21:34He developed cervical lymphadenopathy,
• 21:37and,
• 21:39which was positive for squamous
• 21:41cell carcinoma metastasis.
• 21:44He, rather than having surgery,
• 21:45he underwent radiation to the
• 21:47neck.
• 21:47He then developed,
• 21:49mediastinal metastasis,
• 21:52ultimately was started on it
• 21:54was given targeted radiation to
• 21:56the lung,
• 21:57was trialed on some systemic
• 21:59therapy, but he progressed through
• 22:00disease,
• 22:01and then ultimately,
• 22:03succumbed to his disease.
• 22:05Testing of his tumor revealed
• 22:07class two b.
• 22:08So two very similar appearing,
• 22:11kind of presentations,
• 22:12but the biology of these
• 22:14tumors are dramatically different,
• 22:16and that does get detected
• 22:18on this forty, gene expression
• 22:20profile
• 22:21test. So ways that we
• 22:22can utilize this,
• 22:24for t or 2a and
• 22:262b tumors,
• 22:27would be to consider additional
• 22:29workup to evaluate the nodal
• 22:30basin,
• 22:31CT NET with contrast.
• 22:33You could con also, if
• 22:34you're having any issues with
• 22:35insurance, you could start with
• 22:36an ultrasound,
• 22:38and then assessing the nodal
• 22:39basin with a lymph node
• 22:41biopsy if there is lymphadenopathy
• 22:43just proceeding with, neck dissection.
• 22:45Also, potentially
• 22:47a role for,
• 22:48adjuvant treatment decision making,
• 22:51such as radiation therapy, and
• 22:53doctor,
• 22:54Young will speak about that,
• 22:55and doctor Tran will speak
• 22:57about chemotherapy.
• 22:59And then also the frequency
• 23:00of follow-up for these patients
• 23:02could be increased based on
• 23:04their their,
• 23:05forty GEP results. So an
• 23:07example would be to escalate
• 23:08the care for a high
• 23:10risk,
• 23:10GEP even if the AJCC
• 23:13eighth edition has them as
• 23:14a as a t one.
• 23:15And similarly,
• 23:16consider de escalation. I would
• 23:17do that in the context
• 23:18of a multidisciplinary
• 23:20group,
• 23:20the care for low risk
• 23:22GEP even if it was
• 23:23a, like, a t three.
• 23:25So takeaway points from this,
• 23:27this was this, gene,
• 23:30forty, GEP test was validated,
• 23:34retrospectively.
• 23:35And I would say with,
• 23:37some advanced stage tumors underrepresented,
• 23:40further prospective studies are certainly
• 23:42needed for us to have
• 23:43a better handle on how
• 23:44this can impact, future treatment
• 23:46and patient outcomes.
• 23:48There is a role, I
• 23:49think, for using it as
• 23:50an adjunct to our current
• 23:51guidelines, but not to replace
• 23:53them and to have it
• 23:54be part of a multidisciplinary,
• 23:56discussion.
• 23:57And again, used could be
• 23:58used to escalate care. So
• 24:00that first case presentation,
• 24:02perhaps that I'm sorry, the
• 24:03second case presentation where he
• 24:05had a rapid recurrence,
• 24:06perhaps he's somebody that could
• 24:08have gone quickly to radiation,
• 24:11to,
• 24:12as a response to that
• 24:13T2,
• 24:15or the the 2b class
• 24:17from the forty GEP.
• 24:19And I think it's also
• 24:20important to note that,
• 24:22of the studies included here
• 24:23and a majority of the
• 24:24ones that have been published,
• 24:25they are, industry funded,
• 24:28in some way. So I
• 24:29think it's just important to
• 24:30to keep it to to
• 24:31keep that in mind as
• 24:33we continue to determine the
• 24:35usability of these tests.
• 24:38Thank you.
• 24:43Alright. Thank you, Ansley. I'm
• 24:46going to go ahead and
• 24:47present the, radiation oncologist
• 24:50perspective
• 24:51on,
• 24:52our approaches to cutaneous,
• 24:55squain.
• 24:57Let me just get these
• 24:58slides going.
• 25:00Alright.
• 25:01So,
• 25:02I'm Melissa Young. I'm, the
• 25:04radiation oncologist on the panel
• 25:06today,
• 25:07one of the specialists in
• 25:08our,
• 25:09head and neck,
• 25:11department within the the department
• 25:13of radiation oncology.
• 25:15And I just want really
• 25:15wanted to focus specifically on,
• 25:19you know, how
• 25:20how we look at these
• 25:21high risk skin cancers and
• 25:22how we have to think
• 25:23about the use of adjuvant
• 25:24radiotherapy
• 25:25as a risk reductive strategy
• 25:27to to reduce risk of
• 25:28recurrence.
• 25:29We've already heard the data
• 25:30indicating that these high risk,
• 25:32very high risk cancers that
• 25:33have a, you know,
• 25:35potential risk for recurrence and
• 25:37how are we going to
• 25:38kind of use this data
• 25:39and potentially gene expression profiling
• 25:42to to kind of refine
• 25:43our selection of who may
• 25:45warrant adjuvant treatment and who
• 25:47could potentially,
• 25:49forego adjuvant treatment.
• 25:51So,
• 25:52as has already been demonstrated
• 25:54that the mainstay of treatment
• 25:56for skin cancer,
• 25:57is for surgery upfront.
• 26:00There are situations in where
• 26:01radiotherapy might be used in
• 26:03the definitive setting. It's usually,
• 26:06reserved for patients who are
• 26:07either deemed unresectable
• 26:09or are refusing surgery itself.
• 26:12When it comes to early
• 26:13stage cutaneous,
• 26:15squamous,
• 26:16radiation has equivalent cure rates
• 26:18to that of surgery, but
• 26:19it is time intensive. Radiation
• 26:21is not a single treatment,
• 26:22much like surgery can be
• 26:23done in a single office
• 26:24visit. This is eight sessions
• 26:25over the course of a
• 26:26month at minimum,
• 26:28but many may require thirty
• 26:29to thirty five sessions over
• 26:30six to seven weeks. And
• 26:32there also are going to
• 26:33be, morbidity and side effects
• 26:35associated with radiation treatment.
• 26:38So it's not necessarily going
• 26:39to be our go to
• 26:40for upfront management and it
• 26:41could potentially preclude the use
• 26:43of radiation if the patient
• 26:44were to ever have a
• 26:45second cancer because you really
• 26:47can't, shouldn't give radiation to
• 26:48the same part of the
• 26:49body twice.
• 26:50So
• 26:51therefore radiation is really an
• 26:53adjunct, in the adjuvant setting
• 26:55for high risk or very
• 26:56high risk skin cancers.
• 26:58Now I know that this
• 26:59has already been reviewed to
• 27:01some degree. This is just
• 27:03NCCN guidelines of kind of
• 27:04where radiation
• 27:05therapy may play in
• 27:07in, the
• 27:09care for a patient with
• 27:11a high risk or very
• 27:12high risk skin cancer. And
• 27:13again, just pointing out that
• 27:14this is really at the
• 27:15end of treatment,
• 27:16surgeries performed, you're looking at
• 27:18patients who have positive margins
• 27:20or other pathologic risk factors
• 27:22for those with negative margins
• 27:23that may warrant consideration
• 27:25of radiation therapy.
• 27:28We've already seen the risk
• 27:29stratification,
• 27:31data that's,
• 27:32in the previous presentation, so
• 27:34I won't belabor that. But
• 27:35again, NCCN has,
• 27:38low risk, high risk, and
• 27:39very high risk,
• 27:40and Brigham and Women similarly
• 27:42has
• 27:43similar pathologic features that would
• 27:45also can be considered as
• 27:47risk factors
• 27:48to maybe help guide those
• 27:49patients who are at higher
• 27:50risk for recurrence.
• 27:53When it comes to adjuvant
• 27:54radiation, again, we're really reserving
• 27:56this usually to our very
• 27:57high risk patients or certain
• 27:59high risk situations
• 28:00like recurrent tumors, patients who
• 28:02are immune suppressed and have
• 28:04higher local recurrence risk,
• 28:06or those with T2B or
• 28:08higher tumors.
• 28:10Interestingly,
• 28:10despite how common cutaneous
• 28:13squamous are, the data,
• 28:16prospective data in terms of
• 28:18the efficacy of adjuvant radiation
• 28:19is really lacking.
• 28:21We really much much of
• 28:23the data has just been
• 28:23derived from retrospective reports,
• 28:26and therefore,
• 28:28really truly teasing out who
• 28:29might truly benefit from treatment
• 28:31or not is is not
• 28:33well defined.
• 28:34Similarly,
• 28:35when it comes to radiation,
• 28:37I'm someone who's offering radiation
• 28:39to a certain part of
• 28:40the body, much like a
• 28:41surgeon is gonna remove the
• 28:42area of the tumor.
• 28:44So radiation,
• 28:46can either be,
• 28:47administered just to the primary
• 28:49site where the where the
• 28:50surgery,
• 28:52removed the tumor itself,
• 28:54especially in a clinically node
• 28:55negative patient, although radiation may
• 28:58also include draining lymphatics.
• 29:00And in these retrospective reports,
• 29:02it's very hard to kinda
• 29:03tease out was this primary
• 29:04site radiation, was it primary
• 29:05plus regional radiation, and so
• 29:07kinda how can we also
• 29:08tease out the efficacy based
• 29:10on our target of the
• 29:11radiation is also less clear.
• 29:14I also want to point
• 29:14out that, you know, we
• 29:16don't want to overdo radiation
• 29:18in every patient who comes
• 29:19to our practice because the
• 29:20more of someone you radiate
• 29:22the more side effects you're
• 29:23gonna then kind of subject
• 29:25that patient to that can
• 29:26affect life on quality of
• 29:27life,
• 29:28especially if you're talking about
• 29:29things in the head and
• 29:30neck region where you're talking
• 29:31about salivary gland exposure, dry
• 29:33mouth, lymphedema,
• 29:35scar tissue, and also potentially
• 29:37again eliminating radiation as an
• 29:39option should there be another
• 29:40cancer in this person's lifetime.
• 29:44Kind of to the point
• 29:45of
• 29:46the limited prospective data regarding
• 29:49the efficacy of adjuvant radiation
• 29:50we really don't have much
• 29:51but there
• 29:53just in the last few
• 29:54months another very large retrospective
• 29:56cohort was described with patients
• 29:59with very high risk disease,
• 30:01including multiple institutions where they
• 30:04looked into their databases finding
• 30:06twelve sixty seven patients who
• 30:07met high risk criteria but
• 30:09of note only twelve percent
• 30:11or so are receiving adjuvant
• 30:12radiation treatment.
• 30:14But in this retrospective cohort
• 30:16what we do find is
• 30:17adjuvant radiation does reduce the
• 30:19risk of both local, local
• 30:20regional as well as nodal
• 30:22recurrence by fifty percent. And
• 30:23here are just the Kaplan
• 30:25Meier curves demonstrating that.
• 30:27Importantly, and I think where
• 30:29we, you know, where gene
• 30:30expression profiling might come in,
• 30:32to help sub stratify is
• 30:33is we're not seeing overall
• 30:35survival benefit. We're definitely seeing
• 30:37local control benefit and there's
• 30:39value to that but we're
• 30:39not necessarily seeing a survival
• 30:41benefit. So if we can
• 30:42identify those patients who could
• 30:43potentially have a survival benefit
• 30:45it might also help us
• 30:46prevent that patient who may
• 30:48ultimately succumb as the second
• 30:49case in the previous presentation.
• 30:52So within this this
• 30:54particular retrospective series they did
• 30:56want to look at patients
• 30:57who may have higher risk
• 30:58disease and they kind of
• 30:59sub stratified this by patients
• 31:01who are t2b but with
• 31:02three of the criteria,
• 31:04the T3s
• 31:05as well as those with
• 31:06LVI and they found within
• 31:08that twelve hundred patients, two
• 31:10forty six, that met this
• 31:11higher risk subtype,
• 31:13thirty percent of that population
• 31:15did receive adjuvant treatment.
• 31:17And we're, we're still seeing
• 31:20similarly a fifty percent reduction
• 31:21in local recurrence but certainly
• 31:23that absolute benefit is better
• 31:24because you've now identified a
• 31:26potentially higher risk subgroup so
• 31:27those patients are more likely
• 31:29to benefit. Still not seeing
• 31:30a disease specific,
• 31:32survival benefit, however.
• 31:35This is where, you know,
• 31:37gene expression profiling is is
• 31:39there's a lot of optimism
• 31:40that this could potentially help
• 31:41further identify subgroups that may
• 31:43benefit from radiotherapy.
• 31:46So, you
• 31:47know, in incorporating this in
• 31:49future randomized trials I think
• 31:50will also really be important
• 31:51of trying to help delineate,
• 31:53without bias,
• 31:55how radiation therapy may may
• 31:57be beneficial,
• 31:58in this group of patients.
• 32:01I wanted to
• 32:02present
• 32:04the same group that has
• 32:05this forty gene expression profile
• 32:07panel
• 32:08also looked at their database
• 32:09of the patients comparing those
• 32:11who had radiation and who
• 32:12did not.
• 32:14And they are seeing a
• 32:15benefit
• 32:16or being able to use
• 32:17their class system to help
• 32:19identify subgroups of patients that
• 32:20may actually truly benefit from
• 32:22radiotherapy
• 32:23and those who may have
• 32:23lower risk disease
• 32:25despite classically being considered high
• 32:27risk or very high risk
• 32:29who could potentially
• 32:31omit radiotherapy as part of
• 32:32their care.
• 32:33So
• 32:35specifically they again this is
• 32:37a gene expression profiling
• 32:39looking at in total about
• 32:41four hundred and some patients
• 32:42in this particular
• 32:44cohort of patients that they
• 32:45evaluated who had classically high
• 32:47risk disease or very high
• 32:49risk disease
• 32:50but subclassifying
• 32:51them into the class I,
• 32:52the class IIA or the
• 32:53class IIB based on the
• 32:55gene expression profiling.
• 32:57Again small numbers only about
• 32:58thirteen twelve twelve to thirteen
• 33:00percent of patients actually receiving
• 33:01adjuvant treatment in this group
• 33:02but kind of spread across
• 33:03class 1a, class 2a and
• 33:05class 2b.
• 33:06And when they again look
• 33:07at the Kaplan Meier curves
• 33:09where we're seeing the largest
• 33:10benefit in the left side
• 33:11of the screen here are
• 33:13the class 2b patients. Their
• 33:14risk of metastasis
• 33:17with,
• 33:18without radiotherapy
• 33:20are indicated by that dotted
• 33:21line in red.
• 33:23Whereas those that did receive
• 33:24radiotherapy
• 33:25actually had a dramatic benefit
• 33:27reduction in risk of recurrence.
• 33:28Now these are low numbers
• 33:30but it does you know
• 33:31suggest that we could identify
• 33:32a group that's more likely
• 33:33to benefit from radiotherapy.
• 33:36The right sided panel really
• 33:37just kind of shows this
• 33:39this benefit in in a
• 33:40different graphical way where class
• 33:42one a patients,
• 33:43despite meeting high risk criteria
• 33:46based on gene expression profiling
• 33:48from the small group of
• 33:49patients may not necessarily
• 33:51benefit from adjuvant treatment because
• 33:52their risk seems to already
• 33:54be so low of recurrence
• 33:55already.
• 33:56Whereas those with class two
• 33:57a are those where we
• 33:58might consider radiotherapy.
• 34:00There does seem to be
• 34:01some potential benefit there, but
• 34:03not quite as much of
• 34:04a benefit as what we're
• 34:05seeing in in that small
• 34:06group of two class two
• 34:07b patients.
• 34:09So some of the take
• 34:10home points I take from
• 34:11from the the forty, you
• 34:13know, gene expression panel that's
• 34:15available out there is is
• 34:16that I think we need
• 34:17prospective trials to better understand,
• 34:20you know, how we can
• 34:21identify patients who benefit from
• 34:24this tool to help us
• 34:25better stratify our patients in
• 34:27terms of risk reductive adjuvant
• 34:29strategies.
• 34:30In this retrospective cohort of
• 34:32patients,
• 34:33against small numbers, but it
• 34:35would seem that they have
• 34:37identified
• 34:38a class one gene expression
• 34:40profile that
• 34:41may
• 34:42help
• 34:43determine patients that
• 34:45radiation could be safely omitted,
• 34:47it could help reduce cost
• 34:48of the healthcare system, but
• 34:49also importantly prevent putting someone
• 34:51through side effects that could
• 34:52affect quality of life moving
• 34:53forward because of the side
• 34:54effects radiation can cause.
• 34:58I think what's gonna be
• 34:59an important question to be
• 35:00answered is how do I
• 35:01how do we know what
• 35:02the benefit of radiation is
• 35:03in terms of what part
• 35:04of the body we're treating?
• 35:06So I mentioned earlier, there's
• 35:08no clear guidance in terms
• 35:09of do you treat the
• 35:10tumors tumor bed only or
• 35:11do you treat the tumor
• 35:12bed plus the nodes?
• 35:14I do know that there
• 35:15is interest in a,
• 35:17in,
• 35:18randomized trial to be designed
• 35:19where they would stratify patients
• 35:21based on class two a
• 35:22or class two b,
• 35:24randomizing
• 35:24them to radiation to the
• 35:26tumor bed only versus tumor
• 35:28bed plus nodal radiation, and
• 35:30c, are we seeing a
• 35:31benefit? Because it's not clear
• 35:32if you just radiate the
• 35:33primary site. Maybe we're preventing
• 35:35cancer cells from spreading to
• 35:36the nodes, and and maybe
• 35:37all the benefit is just
• 35:38treating the tumor bed. So
• 35:40so this trial is hoping
• 35:41to to
• 35:42kind of identify
• 35:43where where the benefit comes
• 35:45because if radiation is important
• 35:47as a risk reductive strategy,
• 35:48but we can identify a
• 35:50more localized radiation field that
• 35:52has less side effects for
• 35:53a patient with the same
• 35:54benefit,
• 35:55I I think that's important
• 35:56for us to to understand.
• 35:58And also this prospective trial
• 35:59would theoretically also help confirm
• 36:02in a prospective manner class
• 36:03one patients could safely have
• 36:05radiation omitted.
• 36:07And then long long, you
• 36:08know, big picture is
• 36:10a small number, only about
• 36:12twelve percent of patients based
• 36:13on a lot of these,
• 36:14you know, retrospective studies are
• 36:15actually receiving radiotherapy
• 36:16despite meeting a high risk,
• 36:18very high risk categories.
• 36:21Can this gene expression profiling
• 36:23help, again, identify those patients
• 36:25who do have higher risk,
• 36:26who would have a greater
• 36:27benefit so that we could
• 36:28help prevent morbidity of recurrence?
• 36:30Are those the patients that
• 36:31it's worth putting through side
• 36:33effects and making sure we're
• 36:34kind of funneling into the
• 36:35radiation oncology,
• 36:36you know, adjuvant therapy?
• 36:39And having that shared decision
• 36:40making, discussion with the patient
• 36:42is also really helpful. Having
• 36:43this gene profile class
• 36:45stratification might help a patient
• 36:47make an informed decision.
• 36:49Since we're not seeing a
• 36:50survival benefit necessarily based on
• 36:52the small numbers,
• 36:53It's really coming down to
• 36:55how much do they value
• 36:56preventing a local recurrence potentially.
• 36:58And how we view these
• 37:00gene expression panels and how
• 37:02we incorporate our medical decision
• 37:04making about adjuvant radiation is
• 37:05probably also going to be
• 37:07ultimately influenced
• 37:09by systemic therapy options becoming
• 37:11available.
• 37:11Doctor. Tramm will talk about
• 37:12this kind of at the
• 37:13end summarizing this, but, you
• 37:15know, other systemic therapies that
• 37:17are available, other risk reductive
• 37:18strategies that may have a
• 37:19different side effect profile
• 37:21may also frame how we
• 37:23offer adjuvant treatment even in
• 37:24the gene expression,
• 37:26profiling era.
• 37:31That's wonderful. I don't see
• 37:33any questions in the chat
• 37:34or the Q and A,
• 37:35but just want to remind
• 37:37everyone to please
• 37:38ask us questions.
• 37:40You know, doctor Roche, I
• 37:42was curious, and also doctor
• 37:43Susie,
• 37:44if you had had received
• 37:46and I asked doctor Roche
• 37:47as she's our our PI
• 37:48locally of the NRG Neo
• 37:50adjuvant
• 37:51cutaneous squamous cell carcinoma trial.
• 37:54Would you be more or
• 37:55less likely to enroll a
• 37:57patient that maybe had a
• 37:58smaller
• 38:00tumor that you may not
• 38:01have been as suspicious about
• 38:03if they had a category
• 38:04two a or b on
• 38:06your trial?
• 38:08Yep. You're referring to the
• 38:09neoadjuvant simlopimab trial? That's yep.
• 38:11Yeah.
• 38:12So a smaller tumor, but
• 38:14a higher higher,
• 38:16biological class on the gene
• 38:19panel.
• 38:20I would I you know,
• 38:21it's interesting because we as
• 38:23we're
• 38:24accruing for this trial and
• 38:25seeing how patients do, you
• 38:27know, it overall has a
• 38:28has a very substantial
• 38:32success rate in previous studies,
• 38:34but we are seeing some
• 38:35patients progressing.
• 38:37And so I do wonder,
• 38:38and I've you know, as
• 38:40as we've been talking about
• 38:41this, I've I was thinking,
• 38:43is is this something that
• 38:44could be added to an
• 38:45existing clinical trial,
• 38:48as an additional piece of
• 38:49information to have about these
• 38:50tumors,
• 38:52we have patients that respond
• 38:53very well to suniplumab,
• 38:55and then we're seeing patients
• 38:56that don't respond and that
• 38:57they progress.
• 38:59So this could be, you
• 39:01know, provide some missing information.
• 39:03But having that information to
• 39:05your point,
• 39:06you know, might might raise
• 39:07some questions about maybe they
• 39:09should just go ahead and
• 39:10have surgery,
• 39:11rather than,
• 39:13rather than enroll in a
• 39:14clinical trial.
• 39:16I would also say that,
• 39:19in response to that question
• 39:20about enrollment in the clinical
• 39:22trial is that one of
• 39:23the issues is sometimes we
• 39:24don't have the full staging
• 39:26picture until after they have
• 39:27the MOSE procedure. Right? If
• 39:29the perineural invasion is not
• 39:31captured in the original biopsy,
• 39:33we're not realizing
• 39:34until,
• 39:35their full stage
• 39:37until after they've had surgery.
• 39:38So having that GEP class
• 39:40prior to surgery,
• 39:42can help restratify,
• 39:44when you may not have
• 39:45all of the high risk
• 39:46features identified
• 39:48prior pre op.
• 39:52Okay.
• 39:54So we're gonna shift gears
• 39:56and,
• 39:57go into the magical world
• 39:59of melanoma.
• 40:02But, you know, some of
• 40:03my introduction is really gonna
• 40:05amplify
• 40:06some incredibly
• 40:07important
• 40:08points that have already been
• 40:10made by doctor Young.
• 40:13I'll share my screen.
• 40:21Screen,
• 40:23it's visible, everyone?
• 40:25Okay.
• 40:26So
• 40:28the
• 40:29current landscape for melanoma
• 40:32about how we determine high
• 40:33risk disease is very different
• 40:36than
• 40:37where we are with cutaneous
• 40:38squamous cell. So again, mostly
• 40:41based on clinical pathologic
• 40:42features.
• 40:43And, really importantly, for melanoma,
• 40:46we really routinely use sentinel
• 40:48node biopsy based on the
• 40:50MSLT one
• 40:52trial. And the clinical path
• 40:54features, many of which, you
• 40:55know, similar to what doctor
• 40:57Suzy had begun the conversation
• 40:59with, are are some very,
• 41:00very well validated,
• 41:03nomograms.
• 41:04And some of the recent
• 41:05changes
• 41:06to the
• 41:08NCCN guidelines
• 41:10are incorporating actually using nomogram
• 41:13driven risk factors instead of
• 41:15the classic
• 41:16clinical pathologic
• 41:17features that had driven those
• 41:19decision making
• 41:20processes. But universally,
• 41:22as
• 41:23a as a group, we've
• 41:24agreed that
• 41:25a sentinel lymph node biopsy
• 41:27risk between five to ten
• 41:29percent
• 41:30requires a discussion,
• 41:31and ten or above
• 41:33usually will be typically recommended
• 41:35for a medically fit patient
• 41:38in which the results of
• 41:39the sentinel node are actually
• 41:41going to be impactful.
• 41:42And
• 41:43remember, the sentinel node is
• 41:45what we use for risk
• 41:46prognostication
• 41:47to
• 41:48better
• 41:49delineate
• 41:50where we are. Are we
• 41:51stage one? Are we stage
• 41:53three? Where are we in
• 41:54between? And it really has
• 41:56a great discriminatory
• 41:57value.
• 41:58However, we're always faced with
• 42:00the challenges. You know,
• 42:02our overtreated patient. Right? There's
• 42:04gonna be some patients in
• 42:06melanoma that are gonna be
• 42:07cured by surgery
• 42:09alone. They don't gain anything
• 42:10by watching them closer. They
• 42:12don't gain anything
• 42:13from getting adjuvant therapy,
• 42:16but they bear real financial
• 42:17and emotional cost as well
• 42:19as toxicity.
• 42:21Then we have the other
• 42:22patients that we're missing, the
• 42:24undertreated patients. And, you know,
• 42:26we worry a lot about
• 42:27these in our most common
• 42:28presentation for melanoma, which is
• 42:30our earlier stage patients.
• 42:32We know there's a small
• 42:33percentage of these patients that
• 42:35are recur
• 42:36and
• 42:37they go untreated
• 42:38and that group gets missed.
• 42:40And when we think about
• 42:41it, it becomes an amplified
• 42:43problem. So we have a
• 42:44very small percentage, but that
• 42:45percentage
• 42:47absolutely
• 42:48represents a large group of
• 42:49patients because if, you know,
• 42:51over eighty percent of our
• 42:52patients are going to be
• 42:53early, even a small percentage
• 42:55of that becomes a bigger
• 42:57problem.
• 42:58And, again, our overall goal
• 42:59is we only want to
• 43:00treat patients that are going
• 43:01to relapse. We want to
• 43:03spare anyone that would be
• 43:04cured with surgery. And in
• 43:05order to do this, we
• 43:06have to try to be
• 43:07better
• 43:08at accurate risk stratification.
• 43:12So again, our AJCC
• 43:13staging, it's it's good, but
• 43:15it's not sufficient. We're striving
• 43:17to keep doing that. That's
• 43:18why it's always iterative, why
• 43:20it gets improved.
• 43:21So it does provide some
• 43:22valuable baseline prognostication,
• 43:25but there's a lot of
• 43:25heterogeneity,
• 43:26there's a lot of overlap
• 43:27within these categories that we're
• 43:29working to parse out. And
• 43:31again, that's the question about
• 43:32not only nomograms, but does
• 43:34the GEP testing
• 43:36help us to get a
• 43:37more individualized
• 43:38risk for these patients?
• 43:41Again, that stage 1b2a
• 43:44patient group is really where
• 43:45I think that this is
• 43:46a real unmet need where
• 43:48we may be under treating
• 43:49but also overtreating
• 43:51part of that population,
• 43:53so
• 43:54we really need to be
• 43:55able to distinguish these groups
• 43:58so gene expression profile arrays
• 44:00have been around for almost
• 44:02fifteen years now in melanoma,
• 44:04again much earlier in their
• 44:05inception and use for cutaneous
• 44:07squamous cell And there's three
• 44:09main ones that are commercially
• 44:10available,
• 44:11two available in the States.
• 44:13One is only available in
• 44:15Germany, so the Castle Bioscience
• 44:17is probably
• 44:18the oldest one that had
• 44:19been used. The more recent
• 44:21one is the,
• 44:22the CPGEP,
• 44:24commonly known as the Merlin
• 44:26assay,
• 44:27which will speak in great
• 44:29detail, both myself and Doctor.
• 44:30Christiansen.
• 44:31And then an eleven gene
• 44:33GEP, the Melligenics, and that's
• 44:35used over in Germany.
• 44:37Each of them actually use
• 44:39completely different genes. So within
• 44:41all of these,
• 44:43prognostic and reference genes, they
• 44:45actually do not overlap.
• 44:48Some use clinical
• 44:49pathologic variables, so that's in
• 44:51the DECISION DX as well
• 44:53as the Merlin assay, but
• 44:54not in the melaninics.
• 44:56The output in the DECISION
• 44:59DX, the thirty one GEP,
• 45:01when it includes clinical pathologic,
• 45:03that makes it the I
• 45:04thirty one GEP.
• 45:05That's broken into four different
• 45:07classes versus the other two
• 45:09are really higher low risk
• 45:10binary categories.
• 45:13The
• 45:14thirty one GEP,
• 45:15that was initially developed actually
• 45:17for overall prognosis,
• 45:19but then is now more
• 45:20developed towards giving
• 45:22some guidance for who we
• 45:23should be doing sentinel nodes.
• 45:25Merlin assay was really more
• 45:26developed with sentinel nodes, although
• 45:28there is perhaps some prognostic
• 45:31cation that can be done
• 45:32versus the g e the
• 45:33eleven g e p is
• 45:34really focused on stage two
• 45:37disease
• 45:37and prognostic
• 45:39variables.
• 45:40These can be variably
• 45:42reimbursed. If we look at
• 45:43the cost of a sentinel
• 45:44node, that can range from
• 45:45nine to eleven thousand dollars
• 45:47in the states,
• 45:49and some of these GEP
• 45:50assays can cost, if they're
• 45:52not covered by insurance, anywhere
• 45:54from,
• 45:55three to seven thousand dollars
• 45:57depending upon insurance status.
• 45:59So and right now, the
• 46:01thirty one GEP has not
• 46:02been recommended
• 46:04by any of the major
• 46:05societies.
• 46:06Just this year, the NCCN
• 46:08on the basis of the,
• 46:10Merlin o o one randomized
• 46:12clinical trial,
• 46:13has included this.
• 46:15And there is actually and
• 46:17doctor Tran will talk a
• 46:18little bit more about the
• 46:19nivo mela phase three trial,
• 46:21which is actually doing a
• 46:23identifying the high risk group
• 46:25and actually randomizing
• 46:27those patients to receive either
• 46:28adjuvant nivolumab
• 46:30or placebo
• 46:31for our high risk stage
• 46:33two patients.
• 46:35So looking a little bit
• 46:36into what is the best
• 46:37data that's available
• 46:39to review, so for the
• 46:40thirty one GEP,
• 46:42there was a prospective decide
• 46:43study,
• 46:45again,
• 46:46for what it was looking
• 46:47for. So the negative predictive
• 46:48value means how much can
• 46:50you trust when someone says
• 46:53that this is not gonna
• 46:55be a positive sentinel note,
• 46:56so very, very good negative
• 46:58predictive
• 46:59value in these low risk
• 47:01patients, which you want. There's
• 47:02a low prevalence. We don't
• 47:04wanna have a lot of
• 47:05people that have disease that
• 47:07we didn't detect.
• 47:09Now
• 47:10the problem though is that
• 47:12it not very good at
• 47:13identifying who are the outliers.
• 47:15Right? If everybody in your
• 47:17test, they can identify and
• 47:19I we already said that
• 47:20there are t one a
• 47:21patients that are going to
• 47:23relapse.
• 47:24If your test can't discriminate
• 47:25between that, you know, there
• 47:27there becomes some issue. But,
• 47:29again, if it's negative, that
• 47:31negative predictive value seems to
• 47:33be very good with this.
• 47:36The MERLIN one is the
• 47:38largest prospective
• 47:39GEP trial to date, you
• 47:41know, seventeen hundred
• 47:43patients,
• 47:44looking from, again, T1a through
• 47:46T3.
• 47:48Again, at patients who we're
• 47:49the most worried about and
• 47:51I think that there's very
• 47:52little debate whether or not
• 47:53they should get a sentinel
• 47:54node biopsy,
• 47:56again,
• 47:58it's really not going to
• 47:59be that discriminatory.
• 48:01However, in the group of
• 48:02patients here, these high risk
• 48:04t one a's, t one
• 48:06b's, t two a's, these
• 48:07are our class,
• 48:09clinical stage one b patients.
• 48:11Right?
• 48:12Really good utility
• 48:14that we're seeing here.
• 48:18Very good discriminatory
• 48:20capacity, and and this may
• 48:21be its greatest
• 48:23value.
• 48:24The assay has a good
• 48:25success rate from also getting
• 48:26the FFPBE
• 48:28tissue,
• 48:29and in these important categories
• 48:31here, this clinical 1b, but
• 48:33also in our older patients,
• 48:34where risks of general anesthesia,
• 48:36which you have to have
• 48:37to have a central note,
• 48:38are different.
• 48:39So in these in the
• 48:40patients that are low risk,
• 48:42again, six point six percent
• 48:44versus these patients that are
• 48:45being
• 48:46identified at high risk and
• 48:48where we're actually confirming that
• 48:49that's the case, I think
• 48:51that this is also a
• 48:52group that's pretty powerful.
• 48:57It's been validated
• 48:58independently now. This is a
• 49:00recent data from, a Dutch
• 49:03trial. Again, sensitivity
• 49:04about ninety two point five
• 49:06percent, a good negative
• 49:08predictive value at ninety four
• 49:10percent.
• 49:11Again, specifically
• 49:12at that group where we
• 49:14actually have a real gray
• 49:15area, where we have real
• 49:16focused conversations in the clinic.
• 49:19And, again, we see discrimination
• 49:21between the recurrence free, distant
• 49:23metastasis free, and the melanoma
• 49:25specific survival
• 49:27between these groups.
• 49:28Okay? And, again, the these
• 49:31as you draw them out,
• 49:32these are conversations
• 49:34that you wanna have with
• 49:35your patients,
• 49:36things that help us discriminate
• 49:38about what we're gonna do,
• 49:39what decisions that we're going
• 49:41to make.
• 49:44And and, again, this is,
• 49:47from,
• 49:49Yu et al, from, again,
• 49:50just recently published. Again, looking
• 49:52more specifically at t one
• 49:54a melanomas and graphically
• 49:56over here, you can see
• 49:57that the they're detecting some
• 49:59of these outliers,
• 50:01the very group of people
• 50:02that may have high risk
• 50:03features. Again, it's a low
• 50:05prevalence. It's a low group
• 50:06of patients. So, again, we'll
• 50:08all have to benefit the
• 50:09cost and benefit
• 50:10of that as well as
• 50:11when we think about if
• 50:12it's a low prevalence,
• 50:14you may have more false
• 50:15positives. Right? So a lot
• 50:17of this is focused on,
• 50:18you know, if it's a
• 50:19negative test, can you trust
• 50:20that? Right? So things that
• 50:22we all have to think
• 50:23about. And, again, a lot
• 50:24of this data, you could
• 50:25spend hours delving into.
• 50:28For head and neck populations
• 50:29where we worry about false
• 50:31negatives in sentinel nodes, that
• 50:33that rate is higher than
• 50:34what we have on the
• 50:34trunk and extremities,
• 50:36again, recent study just published,
• 50:38again, looking to validate,
• 50:41the
• 50:42the CPGP,
• 50:43again, seems to be
• 50:45none not different than what
• 50:47we're seeing in the extremities,
• 50:49which is really important.
• 50:53So
• 50:54when we're looking just for
• 50:55sentinel node and decision making,
• 50:57again,
• 50:58if we kind of look
• 50:59at things head to head,
• 51:00there's some benefits but detriments,
• 51:02I think, of each of
• 51:03the assays. The thirty one
• 51:05GEP and I thirty one
• 51:06GEP, which has been around
• 51:07the longest,
• 51:09you know, is not great
• 51:11at discriminating
• 51:12amongst that t one a
• 51:14category,
• 51:15where it hasn't been validated.
• 51:19There's
• 51:20with both of them, we're
• 51:21gonna see that there's decreasing
• 51:23the number of central lymph
• 51:24nodes in theory that we
• 51:25would perform.
• 51:27Almost all of this except
• 51:28for the DECIDE
• 51:29trial was also done only
• 51:31in retrospective
• 51:32series and really wasn't powerfully
• 51:34paired like the MERLIN trial
• 51:36was with that assay.
• 51:39And that really becomes the
• 51:41difference why that
• 51:43got the nod from the
• 51:44NCCN,
• 51:45just the number and the
• 51:47work that went into
• 51:49prospectively
• 51:50following this,
• 51:52internationally
• 51:53for this.
• 51:54Again,
• 51:55some important
• 51:56take homes
• 51:57to always think about with
• 51:59these.
• 52:00There is industry funding and
• 52:01authorship
• 52:02bias. Even the best of
• 52:03us when we when these
• 52:05are funded, there's a lot
• 52:06of input from these,
• 52:09particularly
• 52:10for
• 52:10the g,
• 52:11the, thirty one gene GP.
• 52:15A lot of the data,
• 52:16if you look at it,
• 52:17they've been republishing some of
• 52:18the similar datasets,
• 52:20and that's a practice that
• 52:22one has to be wary
• 52:23about. So it can it
• 52:24can almost
• 52:25the value of that because
• 52:27if you're publishing on the
• 52:28same data and including some
• 52:29and the other, it just
• 52:30becomes really tricky when you're
• 52:32actually trying to go through
• 52:33the literature.
• 52:35Again, the primary objective technically
• 52:37wasn't met in the Merlin
• 52:38o o one,
• 52:40and and that was a
• 52:41higher than expected rate, and
• 52:42the the low risk not
• 52:44keeping that below that five
• 52:46percent threshold that we've tried
• 52:48to, you know, carve out
• 52:50as when we should consider
• 52:51doing it. So seven point
• 52:52one, is a little bit
• 52:54higher than what they what
• 52:55they thought would be.
• 52:58There's no survival benefit. It's
• 53:00not paired yet. Again, a
• 53:02little bit out of the
• 53:02scope of the conversation. Again,
• 53:04we talked about largely retrospective
• 53:06with the exception of the
• 53:07Merlin trial.
• 53:09Again,
• 53:10really needs to discriminate
• 53:12the very groups that we
• 53:13wanna use them. And if
• 53:14someone wants to use a
• 53:15study, the threshold has to
• 53:17be really high and how
• 53:18well this is gonna be
• 53:19to justify the cost that
• 53:21it would potentially
• 53:22be.
• 53:23And I think that, you
• 53:24know, I alluded to it
• 53:25a little bit earlier,
• 53:26there's high pulse
• 53:28false positive rates. And the
• 53:29very first time that I
• 53:30ever saw a patient come
• 53:31into clinic
• 53:33was when I was a
• 53:34fellow in twenty fourteen,
• 53:36and a woman was three
• 53:37months postpartum
• 53:38and had had a one
• 53:40point eight millimeter non ulcerated
• 53:42nodular melanoma with a negative
• 53:44sentinel node and walked into
• 53:46clinic with a class two
• 53:48result
• 53:49on her GEP.
• 53:51And
• 53:52we were
• 53:53at a loss about what
• 53:54do we do.
• 53:56Again, I I thankfully graduated
• 53:58from fellowship, but I I
• 53:59I do think about and
• 54:00I wonder
• 54:01about this young woman who
• 54:03was thirty eight years old
• 54:04who had come into the
• 54:05clinic.
• 54:06And
• 54:07that psychological
• 54:09cost was something that has
• 54:11stuck with me
• 54:12for the last
• 54:14decade
• 54:15about just seeing
• 54:17what the cost is to
• 54:18our patients
• 54:19if they have a false
• 54:20positive. It's nice, you know,
• 54:22to have a really good
• 54:23negative predictive value and say,
• 54:25okay. If it's negative, you're
• 54:26good.
• 54:28But when we're not really
• 54:30capturing or we're causing undue
• 54:32cost to patients, and like
• 54:34I said, it's not just
• 54:35about
• 54:36screening and so forth, it's
• 54:37also about what does it
• 54:38do to them, again,
• 54:40we need to be really,
• 54:41really thoughtful about that. And
• 54:43and, again, just
• 54:44echoing all of my colleagues,
• 54:46you know,
• 54:47GEP, I think it really
• 54:49should inform shared decision making.
• 54:51It doesn't replace
• 54:53the conversations
• 54:54that we have and continued
• 54:56thoughtful interaction
• 54:58or replace the other tools
• 55:00that we have.
• 55:02So with that, I will
• 55:03I will close.
• 55:06Stop my share.
• 55:15Sorry, guys.
• 55:16Can't stop sharing. Oh, there
• 55:18it is.
• 55:19Thanks, Kelly. That was great.
• 55:23So I am up next.
• 55:25Let me see if I
• 55:25can figure out the
• 55:27sharing process here.
• 55:32Okay. Great.
• 55:34So I'm Sean Christiansen. I'm
• 55:36a dermatologist
• 55:37and Mohs surgeon, and I'm
• 55:38gonna shift gears slightly. And
• 55:40instead of giving an overview
• 55:42of GEPs, I'm gonna focus
• 55:44on one
• 55:45specific utilization
• 55:47of GEP,
• 55:48in a clinical trial, which
• 55:49is the ICEFAN trial, which
• 55:50I'll get to talk about,
• 55:52in just a couple minutes.
• 55:54So these are my disclosures,
• 55:55and they're not relevant to
• 55:56what I'm talking about today
• 55:58except for the last one.
• 55:59I I will be an
• 56:00investigator,
• 56:02on the the trial, which
• 56:03is sponsored by Skyline
• 56:05Diagnostics,
• 56:06the company that makes the
• 56:07Merlin
• 56:08GEP.
• 56:10So I'll talk first about
• 56:11principles of surgical management of
• 56:13localized melanoma.
• 56:15I'm gonna take a slight
• 56:16tangent to focus on pathologic
• 56:18margin analysis and why that's
• 56:19important.
• 56:21And then we'll talk about
• 56:22how we can use GEP
• 56:24for predicting nodal metastasis
• 56:25risk and how that influences
• 56:27our decision making and treatment.
• 56:29And then I'll tell you
• 56:30about the the trial,
• 56:32that we're starting up at
• 56:33Yale, hopefully, sometime soon.
• 56:36So surgical management is really
• 56:38the primary treatment for localized
• 56:40and early stage melanoma.
• 56:42And just to review, the
• 56:43goals of surgical excision are
• 56:45are to accurately stage the
• 56:46tumor, which is just as
• 56:48important now as it ever
• 56:49was, especially in the era
• 56:51of targeted therapy and immunotherapy.
• 56:53We we really need to
• 56:54know what the risk factors
• 56:55are to make those decisions.
• 56:57We also wanna optimize local
• 56:59control by achieving complete tumor
• 57:01removal
• 57:02while balancing that with optimizing
• 57:04functional and cosmetic outcomes from
• 57:05surgery.
• 57:06And about fifty years ago,
• 57:08the trend was to really
• 57:09just take as wide of
• 57:10a margin as possible.
• 57:12But since then, we've had
• 57:13some trials that have shown
• 57:14that wider surgical margins are
• 57:16not necessarily better.
• 57:18So a large trial sponsored
• 57:20by the WHO,
• 57:22showed that one centimeter versus
• 57:24three centimeter excision margins for
• 57:26melanoma less than two millimeters
• 57:27in-depth,
• 57:29there was no difference in
• 57:30disease free survival.
• 57:31And then later on in
• 57:32another trial,
• 57:34a deeper melanoma greater than
• 57:36two millimeters in-depth showed that
• 57:37two centimeters
• 57:39was not inferior to four
• 57:40centimeters. So there's a trend
• 57:42for us taking narrower margins,
• 57:43which can still be curative
• 57:45for melanoma.
• 57:46And the question is how
• 57:47narrow can we go?
• 57:49Before we answer that question,
• 57:51we have to recognize
• 57:52that skin cancer often does
• 57:54has have irregular extension.
• 57:56So this is something that
• 57:57came out of a, a
• 57:59publication in Annals of Plastic
• 58:00Surgery many years ago. They
• 58:02actually mapped out the three-dimensional
• 58:05extension of the skin cancer,
• 58:06and you can see how
• 58:07it has many of these
• 58:08irregular
• 58:09strands that penetrate through the
• 58:10subcutaneous tissue.
• 58:12And these irregular and occult
• 58:13islands of tumor may not
• 58:15be detected with standard histologic
• 58:17sections.
• 58:19So this is why the
• 58:20way that we process things
• 58:21is actually really important.
• 58:23So with standard,
• 58:24perpendicular or bread loaf sections,
• 58:27an elliptical specimen like this
• 58:29is cut into these perpendicular
• 58:31slices. Right? So we're only
• 58:32looking at the areas where
• 58:33we happen to take slices.
• 58:35And if there is one
• 58:36focal area that has a
• 58:37positive margin with an irregular
• 58:39extension,
• 58:40that might not be captured
• 58:41on these representative sections.
• 58:43In contrast, if we use
• 58:45en face sections with complete
• 58:46margin analysis, we're looking at
• 58:48the entire peripheral and deep
• 58:50margin.
• 58:50So any small area
• 58:52of positivity at the margin
• 58:54would, in fact, be detected.
• 58:56And this is the principle
• 58:57behind Mohs skin cancer surgery.
• 59:00Just to review this very,
• 59:01very briefly, this is a
• 59:02staged excision of skin cancer
• 59:04with intraoperative
• 59:05frozen sections
• 59:07examining the complete surgical margin,
• 59:09which are read by the
• 59:10surgeon as pathologist.
• 59:12And it's mostly used for
• 59:13basal and squamous cell carcinoma.
• 59:15It has not traditionally been
• 59:16used for melanoma
• 59:18because it has traditionally been
• 59:19more difficult
• 59:20to identify melanoma on frozen
• 59:23sections.
• 59:24But recently, that has slightly
• 59:26changed.
• 59:27So now we can actually
• 59:28use Mart one or melan
• 59:29a immunostains
• 59:31on frozen sections to identify
• 59:32melanocytes.
• 59:33So, these two images here
• 59:35are frozen sections of skin.
• 59:37One of them is a
• 59:38floor example of melanoma in
• 59:39situ. The other one is
• 59:41normal skin with some freeze
• 59:42artifact.
• 59:43Any pathologist in the audience
• 59:44will immediately recognize that it's
• 59:46the one on the left
• 59:47that's melanoma.
• 59:48But for the rest of
• 59:49us, it really helps to
• 59:50have the immunostain.
• 59:51So by using this immunostain,
• 59:53we can clearly see there's
• 59:54confluence, there's nesting, there's upward
• 59:56scatter of melanocytes in the
• 59:57melanoma in situ. But in
• 59:59normal skin, we just see
• 01:00:00a normal distribution.
• 01:00:03And a dual center study
• 01:00:04found that there's very high
• 01:00:05concordance between most sections and
• 01:00:07permanent pathology when using immunostains
• 01:00:09to detect,
• 01:00:10melanoma.
• 01:00:11So, at Yale, we've actually
• 01:00:13been doing this for a
• 01:00:13few years now. Mostly melanoma
• 01:00:15in situ, but a few
• 01:00:16superficially invasive melanoma cases.
• 01:00:19So this is a woman
• 01:00:20with a melanoma in situ
• 01:00:21in a very cosmetically sensitive
• 01:00:23location on the nose. And
• 01:00:24so we wanna try and
• 01:00:25take as narrow of a
• 01:00:26margin as possible. So I
• 01:00:27ended up taking a a
• 01:00:28four millimeter margin around this
• 01:00:30initially.
• 01:00:31And the area in red
• 01:00:32is where I processed a
• 01:00:34deep bulk specimen just to
• 01:00:35look at that and make
• 01:00:36sure there was no invasive
• 01:00:37tumor.
• 01:00:38There was, in fact, just
• 01:00:39melanoma in situ that was
• 01:00:40present.
• 01:00:41But on that first layer,
• 01:00:43there was a positive margin
• 01:00:44at the superior edge only.
• 01:00:46So I then was able
• 01:00:47to take an additional stage
• 01:00:48from there,
• 01:00:50which was clear.
• 01:00:51And this was the patient's
• 01:00:52defect in the end. So
• 01:00:54in the area inferiorly towards
• 01:00:56the alar rim, I was
• 01:00:57actually able to take a
• 01:00:58narrower margin than what is
• 01:00:59recommended by the NCCN.
• 01:01:01And yet in the area
• 01:01:02where the melanoma was extending
• 01:01:03farther superiorly, I was able
• 01:01:05to take a wider margin
• 01:01:06and capture the melanoma to
• 01:01:08achieve clear margins.
• 01:01:10And then we can actually
• 01:01:11do a local flap repair
• 01:01:12and and spare her a
• 01:01:14staged interpolation flap from the
• 01:01:15forehead.
• 01:01:16So so far, we've done
• 01:01:18a hundred and eighty cases,
• 01:01:19of melanoma with Mohs here
• 01:01:21at Yale, and this number
• 01:01:22continues to grow.
• 01:01:24Okay.
• 01:01:25The data supports the use
• 01:01:26of Mohs surgery both for
• 01:01:28melanoma in situ and for
• 01:01:29invasive melanoma.
• 01:01:31So far, our data is
• 01:01:32only retrospective, so we don't
• 01:01:33have prospective data. But the
• 01:01:35retrospective data shows that there
• 01:01:36is a a clear benefit
• 01:01:38in terms of local recurrence
• 01:01:40with Mohs surgery,
• 01:01:41being less than one percent
• 01:01:43compared to standard excision,
• 01:01:44which is about eight percent.
• 01:01:46And if we do staged
• 01:01:47excision, not via Mohs with
• 01:01:49permanent suction processing, but with
• 01:01:51complete margin analysis,
• 01:01:53we can achieve a comparable
• 01:01:55low rate of local recurrence.
• 01:01:56So, again, it's not magic
• 01:01:58of Mohs surgery.
• 01:01:59It's the way that we
• 01:02:00examine the pathologic margin.
• 01:02:02And so we hypothesize
• 01:02:04that excision with complete margin
• 01:02:06analysis may allow for both
• 01:02:08more accurate
• 01:02:09and more complete tumor removal
• 01:02:10with narrower surgical margins and
• 01:02:12thereby decreased surgical morbidity.
• 01:02:16Okay. So we're gonna go
• 01:02:17back to our general principle
• 01:02:19here, of accurately staging a
• 01:02:21tumor, and this is why
• 01:02:21sentinel lymph node biopsy is
• 01:02:23so critical.
• 01:02:24It gives us that information
• 01:02:26on micronodal
• 01:02:27metastases,
• 01:02:28which will influence adjuvant treatment
• 01:02:30decisions.
• 01:02:31However,
• 01:02:32GEPs now offer potentially
• 01:02:35another way to access the
• 01:02:36same question.
• 01:02:38So this is the CPGEP,
• 01:02:40the clinical pathologic gene expression
• 01:02:42profiling
• 01:02:43Merlin test.
• 01:02:44This is a large prospective
• 01:02:47multicenter
• 01:02:47blinded evaluation,
• 01:02:49that doctor Alino already presented.
• 01:02:51Just wanna highlight a couple
• 01:02:52features here.
• 01:02:54When I say blinded, it
• 01:02:55means that the surgeons doing
• 01:02:56the sentinel lymph node procedure
• 01:02:58did not know the patient's
• 01:03:00GEP results, so they were
• 01:03:02blinded to that.
• 01:03:04And, also, just to highlight
• 01:03:05that this GEP includes both
• 01:03:07the gene expression as well
• 01:03:08as the patient's age and
• 01:03:09Breslow death.
• 01:03:11And that's part of the
• 01:03:12reason why for these t
• 01:03:13three tumors,
• 01:03:15a very small fraction of
• 01:03:16them are actually classified as
• 01:03:18low risk,
• 01:03:19both because they have an
• 01:03:21increased,
• 01:03:22sort of aggressiveness of their
• 01:03:23gene expression, but also because
• 01:03:25this test includes the Breslow
• 01:03:27depth. So it's actually pretty
• 01:03:28rare to get a low
• 01:03:29risk result,
• 01:03:31with these t three tumors,
• 01:03:32and we can see that
• 01:03:32it doesn't very accurately predict,
• 01:03:35sentinel lymph node status in
• 01:03:36these groups of patients.
• 01:03:38However, if we go to
• 01:03:39this stage, the stage one
• 01:03:40b subgroup, which is t
• 01:03:42one b and t two
• 01:03:43a, and you'll note that
• 01:03:44this is actually the largest
• 01:03:45fraction of patients in this
• 01:03:46study.
• 01:03:48About half of these patients
• 01:03:49overall were were classified as
• 01:03:51GEP low risk, and we
• 01:03:53see that it really did
• 01:03:54impact the rate of sentinel
• 01:03:56lymph node positivity.
• 01:03:57So in the t one
• 01:03:58b's, if they were low
• 01:03:59risk, they were only five
• 01:04:00percent, sentinel lymph node positive
• 01:04:02compared to sixteen percent in
• 01:04:04the high risk,
• 01:04:06similar for t two a.
• 01:04:07So it's the stage one
• 01:04:08b subgroup where this test
• 01:04:10can be most useful.
• 01:04:11And this is asking
• 01:04:13only the specific question of
• 01:04:15can we predict sentinel lymph
• 01:04:17node status and thereby perhaps
• 01:04:19avoid sentinel lymph node in
• 01:04:20certain patients.
• 01:04:22So then that brings us
• 01:04:23to the trial.
• 01:04:24So this is the ICEMAN
• 01:04:26trial intelligent choice of excision
• 01:04:28margin. And the primary objective
• 01:04:30is to compare three year
• 01:04:32recurrence free survival in patients
• 01:04:34that are treated with narrow
• 01:04:35excision versus wide excision.
• 01:04:37The secondary objective is to,
• 01:04:40assess disease free survival, especially
• 01:04:42looking at nodal disease
• 01:04:45in patients who were treated
• 01:04:46with nodal surveillance without a
• 01:04:48sentinel lymph node biopsy.
• 01:04:49So here's a setup of
• 01:04:50the trial. Patients with clinical
• 01:04:52stage one invasive melanoma
• 01:04:54that are classified as Merlin
• 01:04:55low risk. So all patients,
• 01:04:58are tested with the GEP,
• 01:05:00and only those that are
• 01:05:01low risk are enrolled in
• 01:05:02this study. They're randomized to
• 01:05:04either narrow excision or wide
• 01:05:05excision. And we can see
• 01:05:07what that actually means. Narrow
• 01:05:08excision is zero point five
• 01:05:09centimeters
• 01:05:10with complete margin analysis for
• 01:05:12high risk locations or high
• 01:05:14risk pathology.
• 01:05:15And then the wide excision
• 01:05:16is standard treatment per NCCN
• 01:05:18guidelines.
• 01:05:20And none of these patients
• 01:05:21will have a sentinel lymph
• 01:05:23node biopsy as part of
• 01:05:24the trial.
• 01:05:26And these are the inclusion
• 01:05:27criteria,
• 01:05:28only t one b, t
• 01:05:29two a,
• 01:05:30or
• 01:05:31t one a with depth
• 01:05:33depth greater than zero point
• 01:05:34five and greater than one
• 01:05:35high risk feature, which includes
• 01:05:37lymphovascular
• 01:05:38invasion or mitoses.
• 01:05:40And they all have to
• 01:05:41be GEP low risk.
• 01:05:44In case you need to
• 01:05:45remind yourself what these categories
• 01:05:47are, these are the actual
• 01:05:48preslow deaths.
• 01:05:49And so I mentioned that
• 01:05:50sentinel lymph node biopsy is
• 01:05:52not part of the trial,
• 01:05:53but it may be performed
• 01:05:54at physician discretion, which is
• 01:05:55important.
• 01:05:56So if we feel that
• 01:05:57a patient really would benefit
• 01:05:59from having sentinel lymph node
• 01:06:00biopsy even with a GEP
• 01:06:02low risk score, then then
• 01:06:03we can still do that
• 01:06:04as part of this trial.
• 01:06:06This is a large study
• 01:06:07with the planned enrollment of
• 01:06:08a thousand patients at multiple
• 01:06:10sites. It is currently already
• 01:06:11enrolling in a few sites,
• 01:06:13and, hopefully, Yale will be
• 01:06:14coming on soon.
• 01:06:16For these early stage melanoma,
• 01:06:18we expect a very good
• 01:06:19recurrence free survival of ninety
• 01:06:21eight percent at three years.
• 01:06:22So we really have to
• 01:06:23have a large study to
• 01:06:23be able to detect any
• 01:06:25potential difference here, and this
• 01:06:27is a noninferiority
• 01:06:28design.
• 01:06:29And we're also looking at
• 01:06:30secondary outcomes of surgical morbidity
• 01:06:32and quality of life measures,
• 01:06:34which we expect,
• 01:06:35will be less in patients
• 01:06:37that are randomized to narrow
• 01:06:38excision.
• 01:06:40So just to summarize,
• 01:06:42know, we're thinking about our
• 01:06:44goals of surgical management,
• 01:06:46including both staging the tumor
• 01:06:48and optimizing local control.
• 01:06:50Complete margin analysis may allow
• 01:06:52us to have less morbid
• 01:06:54and yet still complete surgical
• 01:06:55resection with a narrow initial
• 01:06:57margin.
• 01:06:58And then using a GEP
• 01:07:00can maybe help us do
• 01:07:01some of the staging without
• 01:07:02a sentinel lymph node biopsy,
• 01:07:04in select groups of patients,
• 01:07:06especially t one b, t
• 01:07:07two a melanomas,
• 01:07:09that come back as as
• 01:07:11low risk on their GEP
• 01:07:12testing.
• 01:07:13So I will stop there.
• 01:07:16Thank you.
• 01:07:23Thank you so much, doctor
• 01:07:24Christiansen, sir.
• 01:07:26It's nice everyone has already
• 01:07:28kind of laid out the
• 01:07:30groundwork,
• 01:07:31about all these very fancy
• 01:07:33gene expression profiling assays. And
• 01:07:35so my job now is
• 01:07:36trying to
• 01:07:37take all of this new
• 01:07:39technology and how do we
• 01:07:40integrate it best into clinical
• 01:07:42practice to really
• 01:07:44move the needle in terms
• 01:07:45of improving patient outcomes while
• 01:07:46minimizing morbidity and overtreatment.
• 01:07:49And so one of the
• 01:07:50things that I'll talk briefly
• 01:07:51about is kind of the
• 01:07:52difference between how we use
• 01:07:53these assays in melanoma versus
• 01:07:55in cutaneous squamous cell carcinomas.
• 01:07:58And really there are three
• 01:07:59main goals of trying to
• 01:08:01utilize GEPs in clinical decision
• 01:08:03making. One is, you know,
• 01:08:05the question is, can we
• 01:08:06avoid central lymph node biopsies
• 01:08:08in individuals who are at
• 01:08:10low risk, and reduce the
• 01:08:11morbidity of surgical resection?
• 01:08:14You know, how does that
• 01:08:15score really inform whether or
• 01:08:16not the patient needs adjuvant
• 01:08:18therapy to try to minimize
• 01:08:19the stomach toxicities
• 01:08:21for people who don't need
• 01:08:22additional treatment and will have
• 01:08:24a good outcome regardless?
• 01:08:26And then how do we
• 01:08:27utilize this information
• 01:08:29about risk
• 01:08:30to guide surveillance? Do individuals
• 01:08:33who have high risk tumors
• 01:08:34need more surveillance,
• 01:08:36more frequent visits, more frequent
• 01:08:38examinations and scans as opposed
• 01:08:40to those with low risk
• 01:08:41disease?
• 01:08:43So this is just kind
• 01:08:45of comparing again the MERLIN
• 01:08:46and the CASEL bioscience testing.
• 01:08:49I really only focused on
• 01:08:50the assays that are available
• 01:08:52here in the United States.
• 01:08:54But really, as, you know,
• 01:08:56my colleagues have already discussed,
• 01:08:58Merlin has two categories, high
• 01:09:01versus low, whereas Cassell has
• 01:09:03four different categories, 1A, 1B,
• 01:09:052A, 2Bs in melanoma.
• 01:09:09Really the Merlin is kind
• 01:09:10of geared towards looking at
• 01:09:12predicting sentinel lymph node positivity
• 01:09:15to guide decision making and
• 01:09:16lower risk melanomas,
• 01:09:18whereas CASEL really tries to
• 01:09:20address the question about recurrence
• 01:09:22risk and prognostication.
• 01:09:27The real question is these
• 01:09:29tests are very expensive.
• 01:09:30They're
• 01:09:31time consuming. There's a lag
• 01:09:33time in before you get
• 01:09:34the results.
• 01:09:35And given the cost,
• 01:09:38does this information really add
• 01:09:40overall value to the care?
• 01:09:42You know, we talk about
• 01:09:43nomograms and,
• 01:09:46in determining whether or not
• 01:09:47patients have a high risk
• 01:09:48of the sentinel lymph node
• 01:09:49positivity to determine lymph node
• 01:09:51biopsies.
• 01:09:53We have very good long
• 01:09:54term
• 01:09:56AJCC
• 01:09:57melanoma specific recurrence,
• 01:10:00specific survival,
• 01:10:02and all of these latter,
• 01:10:05you know, statistics are free.
• 01:10:06So, what exactly is the
• 01:10:08GEP testing contributing?
• 01:10:10In addition to this, you
• 01:10:12know, we obviously from the
• 01:10:14pros and cons about the
• 01:10:15discussion and how these assays
• 01:10:16were designed and whether geared
• 01:10:18towards addressing,
• 01:10:19there are obvious flaws.
• 01:10:21They're not perfect. No single
• 01:10:23assay is going to yield
• 01:10:24the exact
• 01:10:26data that is required. And
• 01:10:27there's always
• 01:10:29some stage
• 01:10:30or some factor clinical pathologically,
• 01:10:33that may not be well
• 01:10:35represented in the assay.
• 01:10:37So for example, this is
• 01:10:38the Castle Bio and Science
• 01:10:40testing,
• 01:10:41looking at the estimated proportion
• 01:10:42of patients with melanoma
• 01:10:45who are correctly classified as
• 01:10:46high risk. And you can
• 01:10:47see the stage ones,
• 01:10:49did worse,
• 01:10:51in this setting. Whereas if
• 01:10:53you kind of flip this
• 01:10:54and look at the estimated
• 01:10:55proportion of patients without melanoma
• 01:10:57recurrence that were correctly classified
• 01:10:59as low risk,
• 01:11:01again, the stage twos in
• 01:11:02this situation
• 01:11:04did a little bit worse
• 01:11:05based on the assay.
• 01:11:07So I think in general,
• 01:11:09when you kinda summarize these
• 01:11:10different studies that led to,
• 01:11:13the evaluation of these
• 01:11:15clinical assays in in patients,
• 01:11:18really, the consensus is that
• 01:11:20maybe they're not quite ready
• 01:11:22yet for prime time usage.
• 01:11:25The Merlin has,
• 01:11:27a nod in the NCCN
• 01:11:29to consider it,
• 01:11:30but ultimately, it should not
• 01:11:32supersede,
• 01:11:34clinical pathological features, clinical staging,
• 01:11:38and really should be used
• 01:11:39in the context as an
• 01:11:40investigational tool because there's still
• 01:11:42so much that we still
• 01:11:44need to clarify in regards
• 01:11:45to how best to implement
• 01:11:47them, what they ultimately mean,
• 01:11:49what the long term survival,
• 01:11:51recurrence
• 01:11:53risk factors, and overall survival
• 01:11:55is for these individuals.
• 01:11:56So really right now, they're
• 01:11:58still geared towards the, as
• 01:11:59an investigational
• 01:12:01tool
• 01:12:02as part of clinical trials
• 01:12:03and in the research setting.
• 01:12:06This is kind of the
• 01:12:07melanoma specific survival breakdown by
• 01:12:10stage
• 01:12:11at five years and ten
• 01:12:12years. And as you can
• 01:12:14see, really, we try to
• 01:12:15gear therapy, any additional adjuvant
• 01:12:18therapy post surgery
• 01:12:20in those individuals that have
• 01:12:21a higher risk
• 01:12:23of, morbidity, mortality for melanoma.
• 01:12:26And so really it's those
• 01:12:27stage IIBs and beyond even
• 01:12:29including oligometastatic stage four
• 01:12:35trying to decrease
• 01:12:36recurrence, improve overall survival long
• 01:12:39term. But it's really hard
• 01:12:41just given the heterogeneity of
• 01:12:42this population too to really
• 01:12:44decipher
• 01:12:45who really needs additional therapy
• 01:12:47or who are we overtreating
• 01:12:49and inducing potentially permanent toxicities,
• 01:12:52particularly with immune checkpoint inhibitors.
• 01:12:55These days too, we're also
• 01:12:56moving in the neoadjuvant
• 01:12:58space too, and Doctor. Roche
• 01:12:59has an energy
• 01:13:01neoadjuvant study trying to evaluate
• 01:13:03this in a re response
• 01:13:04adaptive,
• 01:13:05system.
• 01:13:07But, really, can we also
• 01:13:09help improve overall survival by
• 01:13:11moving systemic therapy earlier on?
• 01:13:14So where along the spectrum
• 01:13:16really does, GEP therapy,
• 01:13:19play a role?
• 01:13:20So as of right now,
• 01:13:22there really is no strong
• 01:13:24indication,
• 01:13:26in clinical trial that really
• 01:13:28supports
• 01:13:29determining
• 01:13:30or utilizing GEP
• 01:13:31as a sole marker of
• 01:13:33whether or not we decide
• 01:13:34to do adjuvant therapy in
• 01:13:35these individuals.
• 01:13:37Again, there's some good clinical
• 01:13:39trials in progress that unfortunately
• 01:13:40we don't have the data
• 01:13:42yet on. So the NIVOMELA
• 01:13:44trial is currently ongoing,
• 01:13:46is still enrolling,
• 01:13:48but it's prospective randomized phase
• 01:13:50three study. The question here
• 01:13:52is,
• 01:13:53looking at melanogenics
• 01:13:55to see if that can,
• 01:13:57adequately predict
• 01:13:59in those individuals with high
• 01:14:00risk disease
• 01:14:02whether or not utilization of
• 01:14:03adjuvant nivolumab versus placebo improves
• 01:14:06outcomes in individuals with stage
• 01:14:08two melanoma.
• 01:14:09And so something for the
• 01:14:11future that we all should
• 01:14:12pay attention to once that
• 01:14:13data comes out and whether
• 01:14:15or not we can start
• 01:14:16integrating this in a more,
• 01:14:18evidence based fashion into our
• 01:14:21clinical decision making.
• 01:14:23Now the elephant in the
• 01:14:24room that we've not discussed
• 01:14:26yet this entire night is
• 01:14:27how do we factor in
• 01:14:28ctDNA
• 01:14:29too?
• 01:14:31So there's more utilization
• 01:14:33for ctDNA
• 01:14:34in clinical management,
• 01:14:36again, with,
• 01:14:38kind of caveats that this
• 01:14:40is also,
• 01:14:41semi understudied.
• 01:14:43There's a lot of ongoing
• 01:14:44clinical trials looking at this
• 01:14:46and risk stratifying
• 01:14:47individuals
• 01:14:48with ctDNA
• 01:14:50positivity,
• 01:14:51basically detectable circulating tumor DNA
• 01:14:53postoperative,
• 01:14:54and whether those individuals need
• 01:14:56adjuvant therapy versus those who
• 01:14:58clear their ctDNA
• 01:15:00post op.
• 01:15:01One of the best clinical
• 01:15:03trials that has been published
• 01:15:04to date in this setting
• 01:15:05actually occurred in stage two
• 01:15:07colon cancers.
• 01:15:08So this was the dynamic
• 01:15:10trial looking at ctDNA
• 01:15:12to guide the decision on
• 01:15:14whether adjuvant therapy needs to
• 01:15:15be pursued.
• 01:15:17The three year relapse free
• 01:15:19survival was eighty six point
• 01:15:21four percent amongst those with
• 01:15:23ctDNA
• 01:15:24positivity.
• 01:15:26It was really randomized to
• 01:15:27investigators' choice use that to
• 01:15:29guide decision making whether or
• 01:15:31not they received adjuvant chemotherapy.
• 01:15:33Again, the regimen,
• 01:15:35deferred to the investigator.
• 01:15:37This was versus ninety two
• 01:15:39point five percent
• 01:15:40of
• 01:15:41recurrence free survival amongst the
• 01:15:44CTD negative patients who did
• 01:15:46not receive any additional therapy.
• 01:15:48So in the end,
• 01:15:49based on the CTD guided
• 01:15:51group, fewer patients,
• 01:15:53had to undergo and receive
• 01:15:55adjuvant chemotherapy,
• 01:15:57and overall responses were very
• 01:16:00similar.
• 01:16:01Now the COMBOAD
• 01:16:04trial was utilized to look
• 01:16:06at adjuvant dibrafnib and trametna
• 01:16:08for individuals with BRAF mutated
• 01:16:10melanomas
• 01:16:12for a year in the
• 01:16:13adjuvant setting.
• 01:16:14So what they did in
• 01:16:15this study was actually also
• 01:16:17evaluate, and they reported this
• 01:16:19separately,
• 01:16:20biomarker analysis,
• 01:16:22looking at ctDNA post resection.
• 01:16:24They found that in those
• 01:16:26individuals,
• 01:16:27post resection, thirteen percent of
• 01:16:29patients had still detectable ctDNA,
• 01:16:32and those were associated with
• 01:16:34worse survival outcomes.
• 01:16:36And so, again, ctDNA
• 01:16:38has been utilized
• 01:16:39and is emerging as prognostic
• 01:16:41outcome
• 01:16:42in individuals
• 01:16:44with,
• 01:16:45you know, high risk disease.
• 01:16:47Ultimately,
• 01:16:49if one
• 01:16:50assay
• 01:16:51ctDNA versus GEP is going
• 01:16:53to be superior, inferior versus
• 01:16:53others yet to be determined.
• 01:16:53And I think really
• 01:16:54given
• 01:16:57determined. And I think really
• 01:16:59given kind of the limitations
• 01:17:00on some of the GEP
• 01:17:01assays and what they're designed
• 01:17:03to question whether or not
• 01:17:04that's sentinel lymph node biopsy
• 01:17:06versus overall prognostication,
• 01:17:10where that is aligned also
• 01:17:12kind of depends on the
• 01:17:13assay in question two.
• 01:17:16Whether or not GEP impacts
• 01:17:18how we surveil melanomas,
• 01:17:20also we're still in the
• 01:17:21primitive era of this as
• 01:17:23well.
• 01:17:24Really the guidelines and the
• 01:17:26consensus
• 01:17:27from individuals is that no
• 01:17:28changes really should be made
• 01:17:30at this point
• 01:17:31based on GEP,
• 01:17:33high risk versus low risk
• 01:17:35expression.
• 01:17:39Pathologic staging and histologic
• 01:17:41criteria.
• 01:17:42And that should always still
• 01:17:44take precedence over GEP as
• 01:17:46we're trying to learn more
• 01:17:47and gain more data, particularly
• 01:17:49prospective clinical trial data to
• 01:17:51see whether or not long
• 01:17:53term outcomes,
• 01:17:54hold true.
• 01:17:56Now kind of switching gears
• 01:17:58to the squamous cell carcinoma
• 01:18:00GEP space, so this is
• 01:18:01the CASEL Decision DX for
• 01:18:03squamous cell carcinomas.
• 01:18:05Now unlike melanomas,
• 01:18:07the testing
• 01:18:08of in this particular disease
• 01:18:11has is more mature compared
• 01:18:13to in melanoma.
• 01:18:15So this is the forty
• 01:18:16GEP panel.
• 01:18:18Again, as,
• 01:18:20my prior colleagues already has
• 01:18:21talked kind of three different
• 01:18:23categories, which help to predict
• 01:18:26metastasis free survival.
• 01:18:28Really kind of the three
• 01:18:30year metastasis
• 01:18:31free survival
• 01:18:32is much better in those
• 01:18:34with class one
• 01:18:35at ninety one percent versus
• 01:18:37class IIBs, which are about
• 01:18:38forty four percent.
• 01:18:40And again,
• 01:18:42as Doctor. Roche had mentioned,
• 01:18:44kind of the positive predictive
• 01:18:45value shows improvement over our
• 01:18:47standard
• 01:18:49AGCC
• 01:18:50staging and,
• 01:18:51Brigham and Women's Hospital staging
• 01:18:53systems at this point.
• 01:18:55So again, it was nice
• 01:18:56to see that comparison of
• 01:18:58GEP,
• 01:19:00versus traditional staging algorithms.
• 01:19:03Now the question here also
• 01:19:04is, you know, how do
• 01:19:05we utilize GEP, the Castle
• 01:19:08Bion Science testing for squamous
• 01:19:10cell carcinomas to determine whether
• 01:19:12or not individuals need adjuvant
• 01:19:13therapy?
• 01:19:15There was actually two recent
• 01:19:17updates
• 01:19:18over the past year,
• 01:19:20C POST and keynote six
• 01:19:22thirty, looking at adjuvant anti
• 01:19:23PD-one therapy in these high
• 01:19:25risk
• 01:19:26individuals. So these are high
• 01:19:27risk generally by the MCCN
• 01:19:28criteria. The one utilized for
• 01:19:28C POST is kind of
• 01:19:30listed
• 01:19:39evasion, in transit metastases,
• 01:19:41recurrence, etcetera.
• 01:19:44And so really for C
• 01:19:45post trial,
• 01:19:46they did not report any
• 01:19:47GEP utilization
• 01:19:49nor did keynote sixthirty.
• 01:19:51So again, in this phase,
• 01:19:52it's kind of hard to
• 01:19:53compare the utility of adjuvant
• 01:19:55therapy,
• 01:19:56at this time.
• 01:19:58Now when we look at
• 01:19:59surveillance,
• 01:20:01we know that in terms
• 01:20:02of metastases, the class I
• 01:20:04patients had less than ten
• 01:20:05percent risk of metastases versus
• 01:20:07the class IIb. Those individuals
• 01:20:10had a greater than fifty
• 01:20:11percent risk of metastases.
• 01:20:14So when you have an
• 01:20:15AJCC
• 01:20:16that's a T3, T4
• 01:20:19or a similar kind of
• 01:20:21T2B,
• 01:20:22T3 and the, BHW
• 01:20:24classification,
• 01:20:26those are always kinda intermediate
• 01:20:28or high risk. It's really
• 01:20:30kinda discerning the T one,
• 01:20:32T twos and the T
• 01:20:34one, T two As,
• 01:20:36that really the class one,
• 01:20:38class two A, class two
• 01:20:39B designation kinda really helps
• 01:20:42to better delineate with granularity
• 01:20:45those individuals that will be
• 01:20:46at higher risk for developing
• 01:20:48metastatic disease in the future.
• 01:20:50And so really, it's kind
• 01:20:52of in those individuals where
• 01:20:54the discussion about
• 01:20:55how do you improve upon
• 01:20:58local disease control with the
• 01:21:00addition of adjuvant radiation,
• 01:21:02as Doctor. Young had alluded
• 01:21:03to, that really kind of
• 01:21:04makes
• 01:21:05a difference in long term
• 01:21:07clinical responses.
• 01:21:10So when we think about,
• 01:21:12surveillance in cutaneous squamous cell
• 01:21:14carcinomas too,
• 01:21:16this was kind of an
• 01:21:17interesting paradigm that was presented
• 01:21:19by Farberg et al,
• 01:21:21looking at those individuals that
• 01:21:23are classically considered high risk
• 01:21:25for cutaneous squamous cell carcinoma.
• 01:21:27So considering and taking into
• 01:21:29factor the forty GEP expression,
• 01:21:32and then looking at class
• 01:21:34one individuals. In these,
• 01:21:36class one groups, we know
• 01:21:38that these patients do well
• 01:21:39overall
• 01:21:40based on,
• 01:21:43kind of their three year
• 01:21:44relapse free survival outcomes.
• 01:21:46So if you have someone
• 01:21:48who has is a class
• 01:21:50one, potentially fifty to sixty
• 01:21:52percent of those individuals
• 01:21:54could be spared,
• 01:21:55additional adjuvant anti PD one
• 01:21:58therapy.
• 01:21:59So you're minimizing toxicity.
• 01:22:01You know these patients are
• 01:22:02gonna do well overall.
• 01:22:04And so there is more
• 01:22:06equipoise than to say, okay,
• 01:22:08maybe these individuals don't need
• 01:22:10quite as much surveillance,
• 01:22:12and don't need standard, scheduled
• 01:22:15imaging.
• 01:22:16In those individuals then on
• 01:22:17the opposite side who have
• 01:22:19a class 2b signature,
• 01:22:20those are the ones that
• 01:22:21you really wanna focus on
• 01:22:23and maximizing,
• 01:22:24all the therapies
• 01:22:26upfront,
• 01:22:27to try to minimize,
• 01:22:29distant metastasis,
• 01:22:31recurrence.
• 01:22:32And so for these individuals,
• 01:22:34you wanna consider adjuvant radiation.
• 01:22:37And this is what Doctor.
• 01:22:38Young had already alluded to
• 01:22:40too with these individuals,
• 01:22:41who had adjuvant radiation.
• 01:22:43Those individuals had a fifty
• 01:22:45percent improvement in five year
• 01:22:47metastasis free survival.
• 01:22:49And again, this was something
• 01:22:50that was not seen in
• 01:22:51class one or the class
• 01:22:522a
• 01:22:53GPs.
• 01:22:54The moderate intensity, obviously,
• 01:22:57we consider,
• 01:22:59doing all the adjuvant treatments,
• 01:23:01the adjuvant radiation,
• 01:23:03but at that point too,
• 01:23:04it's all you're just also
• 01:23:06equicosed to talk about, kind
• 01:23:08of risk factors and on
• 01:23:09a patient per case basis,
• 01:23:12you know, what are their
• 01:23:13goals, what are their morbidities
• 01:23:14and trying to factor this
• 01:23:16in on a more personalized
• 01:23:18approach.
• 01:23:19So again, kind of the
• 01:23:21summary of all of this
• 01:23:22is that it's complicated.
• 01:23:24There's still a lot that
• 01:23:25we still have yet to
• 01:23:26learn in trying to integrate
• 01:23:29these new technologies in our
• 01:23:31clinical practice,
• 01:23:32a lot of clinical trials
• 01:23:34that are forthcoming that may
• 01:23:35help to elucidate these,
• 01:23:37questions and answers.
• 01:23:40But really,
• 01:23:41and, you know, trying to
• 01:23:43figure out how do we
• 01:23:45also then reconcile ctDNA data
• 01:23:47with GEP data to best
• 01:23:49guide decision making.
• 01:23:52You know, there may be
• 01:23:53forthcoming
• 01:23:55GEP
• 01:23:56like two point zero at
• 01:23:57some point
• 01:23:58that will also help not
• 01:23:59only look at metastases, local
• 01:24:01recurrence, lymph node positivity,
• 01:24:04but overall survival too.
• 01:24:06And so I know there's
• 01:24:07individuals already working on this
• 01:24:09and emerging
• 01:24:10technologies in this regard too.
• 01:24:14With that, I think
• 01:24:15we are done.
• 01:24:17So,
• 01:24:18let me stop sharing, and
• 01:24:20maybe we can go back
• 01:24:21to the chat if anyone
• 01:24:22has any questions.
• 01:24:46Okay.
• 01:24:49We had,
• 01:24:51Barb Barbara Johnson had a
• 01:24:52question about how long the
• 01:24:54assays take to
• 01:24:56result to,
• 01:24:57doctor Christiansen.
• 01:24:59Do you have a
• 01:25:00you've been using this, I
• 01:25:01think, a little bit more.
• 01:25:03Yeah. So it's it's usually
• 01:25:04about a week
• 01:25:05turnaround time, and most of
• 01:25:07that is from shipping. So
• 01:25:10the the assay is done
• 01:25:12on formalum fixed slides from
• 01:25:14the original biopsy,
• 01:25:16and that has to be
• 01:25:17shipped to the company. And
• 01:25:18then once they get it,
• 01:25:19then they can do that
• 01:25:20test within a day or
• 01:25:21so.
• 01:25:22So it's it's just the
• 01:25:24logistics of transferring it that
• 01:25:25takes a little bit of
• 01:25:26time. But since that's done
• 01:25:28on the original diagnostic biopsy,
• 01:25:30that can actually be initiated,
• 01:25:33you know, while the patient
• 01:25:34is getting set up for
• 01:25:35their consultation and surgery.
• 01:25:41I have a question for
• 01:25:42doctor Christiansen maybe.
• 01:25:45Sure. In terms of these
• 01:25:47various assays, there's so many
• 01:25:49companies that offer their own
• 01:25:51proprietary
• 01:25:52assay.
• 01:25:54Sometimes tissue is quite precious.
• 01:25:56So if you commit to
• 01:25:57doing,
• 01:25:59you know, formalin
• 01:26:00submitting formalin fixed tissue to
• 01:26:03a company for one particular
• 01:26:04assay, for example, we do
• 01:26:06a lot of Natera ctDNA
• 01:26:08monitoring.
• 01:26:10Do any of these companies
• 01:26:11ever work together
• 01:26:13to
• 01:26:14kind of maximize
• 01:26:16the DNA and
• 01:26:18the kind of the RNA
• 01:26:20so that they get as
• 01:26:22much
• 01:26:22out of such little tissue,
• 01:26:24particularly if we're talking about
• 01:26:25trying to prioritize looking at
• 01:26:27comparing ctDNA
• 01:26:29gene expression profiling in some
• 01:26:30of these t one, t
• 01:26:32two tumors where there's not
• 01:26:33gonna be much to begin
• 01:26:34with.
• 01:26:37So I don't think I
• 01:26:37know enough to accurately,
• 01:26:39fully answer that question,
• 01:26:41but my guess is probably
• 01:26:43the answer is no. I
• 01:26:44think these companies are all
• 01:26:46independent,
• 01:26:47and they all do their
• 01:26:48own thing. I mean, I
• 01:26:49think, you know, you you
• 01:26:50presented the idea of using
• 01:26:52GEP GEP and ctDNA
• 01:26:54together in a prospective trial.
• 01:26:55And if that were going
• 01:26:57to happen, then, you know,
• 01:26:58they would have to work
• 01:26:59together.
• 01:27:00But I think short of
• 01:27:01that right now, everybody's independent.
• 01:27:05I I believe that's that's
• 01:27:07correct. And and, actually, I
• 01:27:08believe that the Signatera slash
• 01:27:10Natera company is working on
• 01:27:12their own
• 01:27:13GEP.
• 01:27:15So,
• 01:27:16and and to the point
• 01:27:17made earlier,
• 01:27:18each one is proprietary. So
• 01:27:20there's a reason that the
• 01:27:21genes of interest don't overlap
• 01:27:24with many of these even
• 01:27:26though they're looking at
• 01:27:28melanoma, which I always find
• 01:27:30fascinating.
• 01:27:31And you'll, unfortunately, I don't
• 01:27:33know if we'll we'll ever
• 01:27:34be able to get head
• 01:27:35to head,
• 01:27:36comparisons
• 01:27:37for these because of the
• 01:27:39reason that you said our
• 01:27:40melanomas are small and the
• 01:27:42samples are precious. And,
• 01:27:44you know,
• 01:27:47it it'll be very, very
• 01:27:48interesting. But, again, we never
• 01:27:50thought companies would cooperate for
• 01:27:52systemic immune therapy and look
• 01:27:54where we are now. So
• 01:27:56I'm I'm hopeful.
• 01:28:01Maybe the sponsor of Iceman
• 01:28:03would like to do c
• 01:28:04two DNA.
• 01:28:07They might.
• 01:28:10We'll bring it up at
• 01:28:11the next investigators meeting.
• 01:28:15Alright. Well, for all of
• 01:28:16those folks, I can't believe
• 01:28:18how many people have stayed
• 01:28:19on the entire
• 01:28:20time. So thank you so
• 01:28:21much to your audience. Thank
• 01:28:23you for questions, and thank
• 01:28:25you to my colleagues that
• 01:28:27I have the pleasure
• 01:28:28to work with every day
• 01:28:30and pick their brain about
• 01:28:31just about anything.
• 01:28:33So thank you to everyone
• 01:28:35and, we're all open, very
• 01:28:37easy to find via email.
• 01:28:39If people had questions that
• 01:28:41they come up with, even
• 01:28:42for those of you who
• 01:28:43may be later viewing this
• 01:28:45on YouTube. So thank you
• 01:28:46and have a wonderful evening.