10/14/2020: James McPartland, PhD

Transcript

• 00:00I think we'll get started.
• 00:02Welcome everybody to the
• 00:04Department of Pediatrics.
• 00:06Grand rounds.
• 00:08True if you can advance,
• 00:10and just to let everybody know,
• 00:13we've got some great grand
• 00:15rounds coming up our next grand
• 00:17rounds next week will be Jason
• 00:19Schwartz and very timely childhood
• 00:22vaccination policy during COVID-19.
• 00:24And then he'll be followed by John Hughes,
• 00:27Professor of Medicine,
• 00:29who is actually requested
• 00:30by multiple providers,
• 00:32and he's going to address the
• 00:34rationing of care.
• 00:36An equity with a focus to a case
• 00:38that we had in Pediatrics recently.
• 00:46Anne, I'm very proud to
• 00:48announce our inaugural award for
• 00:51innovation in pediatric education.
• 00:53An our award ES who put a lot of
• 00:57work into their grant applications.
• 01:01Daniela Hochreiter, Julie Leviter,
• 01:03Leslie Harris, Michael Goldman,
• 01:05Ann Shannon O'Malley. We wish them.
• 01:09Congratulations will be hearing more
• 01:12about their curricula and I also big.
• 01:15Thank you to mark our back,
• 01:18Melissa Lang Han and Tom Murray
• 01:21for pioneering this process.
• 01:23So again, congratulations,
• 01:24we're very proud of this award
• 01:27and thanked after Bogue well Cliff
• 01:30for supporting its next slide.
• 01:34So just remind everybody that there's no
• 01:37commercial support for this grand rounds,
• 01:39nor are they any conflicts of interest.
• 01:46And trip to receive your see me or your
• 01:49continuing nursing education credit.
• 01:51You just need to sign into the zoom
• 01:54meeting for the questions you can write
• 01:56them in the chat or the Q&A section,
• 02:00or you can raise your hands out.
• 02:02We're hoping to actually unmute
• 02:04some of you and next it we can.
• 02:09So just to introduce our speaker,
• 02:12Jayme Mcpartland, we have Linda Mays.
• 02:14So, Linda thank you.
• 02:15Thank you very much and so great to
• 02:19actually have another grand rounds in the
• 02:21Department of Pediatrics
• 02:22that brings the child,
• 02:24study center and Pediatrics together.
• 02:26The more we do this, the better.
• 02:28It is definitely a priority for Cliff and me,
• 02:32but it's a special pleasure to be able to
• 02:35introduce my colleagues jayme Mcpartland,
• 02:37whom as you see on the slide,
• 02:40is the associate professor.
• 02:42In the child study center.
• 02:44I'm just very briefly
• 02:45because I really want you.
• 02:46I want you to hear how
• 02:48outstanding a scientist Jamie is,
• 02:50but I really want you to hear that from him.
• 02:53Jamie got his undergraduate degree
• 02:55from Harvard and did his graduate
• 02:58work at the University of Washington
• 03:00with some of the Preeminent Autism
• 03:02investigators of that of the time.
• 03:05In 2004 he came to the child
• 03:07study center are great fortune,
• 03:09initially as an as an intern,
• 03:11and then progressed and is now a member.
• 03:14Have been a member of the faculty and
• 03:17associate professor and really one of our
• 03:20most creative and productive scientist.
• 03:22Jamie's work is very much around the
• 03:25developmental continuum of autism,
• 03:27developmental disabilities
• 03:28and most importantly,
• 03:29understanding the biomarkers how we can
• 03:32early on on detect markers for autism
• 03:35and allow us not only to intervene early,
• 03:39but understand some of the mechanisms of
• 03:42expression of autism across development.
• 03:44He is he,
• 03:45as the Pi of a very very large multi
• 03:49site trial investigating biomarkers and
• 03:52autism that you will hear more about.
• 03:56Matt Grant was just recently renewed
• 03:58and also has been working with
• 04:00the FDA around his biomarker work
• 04:03on this is just a small portion
• 04:05of how creative Jamie is.
• 04:07Also want you to to know that he is
• 04:10clinically immersed in the director of
• 04:12our autism clinical efforts as well.
• 04:15So he definitely is a translation of
• 04:17scientist in every aspect of the word.
• 04:20I'm so.
• 04:22There's much more I could say about Jamie,
• 04:24but I want you to hear from Jamie Ann.
• 04:26I want to turn this now to Jamie
• 04:28for his presentation.
• 04:29Thank you for doing this, Jamie.
• 04:33Thank you so much Linda.
• 04:34It's so nice to have the opportunity
• 04:36to speak with you all today and it's
• 04:38so nice to have the opportunity
• 04:39to be introduced by Linda,
• 04:41who is who knows my work and me so
• 04:43well and who has been such a actually
• 04:46a contributor to some of the work
• 04:48that I'll talk about today are my
• 04:50slides showing as they should.
• 04:52Yes, OK, great. So.
• 04:53As Linda mentioned,
• 04:54I'm going to talk some today about
• 04:56a project that I've LED called
• 04:59the autism biomarkers consortium
• 05:00for clinical trials,
• 05:02and a lot of the work that has fed into it.
• 05:05It's also need to talk about this at
• 05:08this time to a group at Yale and that
• 05:12Linda mentioned that I came here in 2004,
• 05:14and as you'll see,
• 05:16that was the year one of the first
• 05:18studies that I conducted was published.
• 05:21And really, the.
• 05:22The study that launched the line of inquiry
• 05:25that I'll be talking a lot about today.
• 05:29These are my disclosures and I don't
• 05:31think there's any conflicts with the
• 05:32things that I'll talk about today.
• 05:34So this is broadly one I'd
• 05:36like to speak about.
• 05:38I want to talk a little bit about
• 05:40biomarker research in autism,
• 05:42where things stand.
• 05:43What are the what I see as
• 05:45some of the key objectives,
• 05:47and then also some of the key
• 05:50goals targets if you will,
• 05:52both in terms of the science of Biomarkers,
• 05:54but also some of the practical
• 05:56considerations for implementation,
• 05:58I'm going to talk about progress to date.
• 06:00Really focusing on a particular biomarker,
• 06:03an EG biomarker that.
• 06:04We've done a lot of work within
• 06:06my lab and talk about,
• 06:08despite the strides that we've
• 06:10made some of the challenges that
• 06:12remain and then talk about some
• 06:14of the things that I think will
• 06:16be most helpful moving forward in
• 06:18terms of conducting more rigorous
• 06:20studies like the EBCT,
• 06:21but then also thinking about ways,
• 06:23ways that we may be able to
• 06:26more directly in applied ways,
• 06:27use these biomarkers to inform care.
• 06:31So I want to start as Linda mentioned,
• 06:34I'm a I'm a scientist and I'm also
• 06:37a clinician and I want to start
• 06:39just by providing a contrast of the
• 06:42tools that I have at my disposal.
• 06:44In both of these contexts.
• 06:46So when I'm in my lab,
• 06:48we use many different kinds of
• 06:51tools to understand autism.
• 06:52We use e.g I'll talk a lot about EG.
• 06:55Today we collaborate with outstanding
• 06:57colleagues to use methods like
• 06:59positron emission tomography or
• 07:00functional near infrared spectroscopy.
• 07:02And then we do things like eye tracking
• 07:05to understand the way children adults
• 07:07with autism take in the world visually.
• 07:10So what powerful?
• 07:11What an impressive array of tools
• 07:13that I can use as a scientist.
• 07:15But when I go to the clinic and I
• 07:18think about the tools that I have at
• 07:21my disposal to diagnose to think about
• 07:24what treatments are going to be most
• 07:27helpful for a person with autism,
• 07:29have one tool, and it's the same
• 07:31tool that Lee O'Connor used when he
• 07:34first wrote about autism in 1943.
• 07:37I have my my clinical I and we have
• 07:39many wonderful clinicians at Yale
• 07:42and there's many things that can
• 07:45be done with a clinical I alone.
• 07:48But there's many shortcomings
• 07:49as I'll tell you about and.
• 07:52For that reason we want to have
• 07:54methods that are more objective
• 07:56that are more deployable that are
• 07:59biologically based and what we
• 08:02would call these are biomarkers.
• 08:04So biomarker, commune,
• 08:05lots of things that the FDA.
• 08:07Foundation for the National
• 08:09Institute of health have brought
• 08:11together groups to really create
• 08:13working definitions of Biomarkers,
• 08:14and these are the definitions
• 08:16that we adhere to,
• 08:18and so a biomarker is a characteristic
• 08:20that is measured as an indicator
• 08:22of normal biological processes,
• 08:24pathogenic processes or responses
• 08:25to an intervention including
• 08:27therapeutic interventions.
• 08:28So as you can see,
• 08:29it's it's a very broad definition.
• 08:32What's important is that it's
• 08:34something that is rooted in biology.
• 08:36And that can be measured objectively.
• 08:39And this to date we do not have any
• 08:41that we can use effectively in clinical
• 08:44practice or in clinical research in autism.
• 08:47And that really is the thrust of the
• 08:51research that we do in our group.
• 08:53Try to create biological tools
• 08:55for any of these purposes that can
• 08:58improve diagnosis that can help
• 08:59us pick which treatments that are
• 09:02working that to apply and to measure
• 09:04whether treatments are working.
• 09:07There are many different kinds of Biomarkers.
• 09:09I'll tell you just about a few
• 09:12if you're interested in this.
• 09:13You can consult that Working Group Document,
• 09:15which should tell you everything
• 09:18you would ever want to know.
• 09:21A prognostic biomarker, for example,
• 09:22is one that tells us something
• 09:24about the course of disease.
• 09:26For example,
• 09:27a prognostic biomarker electes might
• 09:29tell me what a persons linguistics
• 09:32outcomes would be in the future.
• 09:34A pharmacodynamic response biomarker
• 09:35might tell us whether a targeted system,
• 09:38in our case I tend to think
• 09:40of Neural Systems,
• 09:41is being engaged by a treatment.
• 09:43This is something that really is
• 09:45pretty far off in autism that we're
• 09:47still trying to establish what neural
• 09:50systems we should be targeting.
• 09:52And then Lastly is a diagnostic biomarker,
• 09:54and this tends to be what most people
• 09:57think of when they think about about.
• 10:00Biomarker for autism.
• 10:01The idea that I can apply some kind
• 10:04of biological assay and determine
• 10:07who has autism and who doesn't
• 10:09now that would be a diagnostic
• 10:12biomarker of the condition.
• 10:13Thinking about a biomarker that
• 10:15Maps on to a diagnostic category.
• 10:18This is something that I feel is is
• 10:20a very high bar and a potentially
• 10:23unrealistic bar in that autism
• 10:25isn't a biological entity.
• 10:27Autism is a description of a set of features.
• 10:31That are reflective of many
• 10:33different ideologies and so.
• 10:35This is this is a worthy goal.
• 10:38It's not one that we're so actively
• 10:41pursuing well we are pursuing is
• 10:43to try to find a different kind
• 10:45of diagnostic biomarker,
• 10:47a biomarker that is diagnostic of a
• 10:50subgroup or of stratifications within
• 10:52the group of people who have autism.
• 10:55It's a tremendously heterogeneous group,
• 10:57and so maybe we can find ways of parsing
• 11:00them into subgroups that are meaningful
• 11:03in terms of ideology in terms of outcome.
• 11:06In terms of appropriate treatment,
• 11:08and these are the kinds of biomarkers that
• 11:10I'm going to talk really about ladies,
• 11:13stratification biomarkers,
• 11:13or if we were to think in terms of
• 11:16the some of the more current FDA
• 11:18jargon we might call them diagnostic
• 11:20enrichment Bar Biomarkers.
• 11:22Thinking that these are a stratum that
• 11:24we could define that would could be used
• 11:27to enrich clinical trials for autism.
• 11:29Kind of a distilling to autism
• 11:31to its essence.
• 11:32If you will let me talk about
• 11:33some of the Scientifical's,
• 11:35these are the and.
• 11:36Keep in mind I'm going to follow on
• 11:38these and tell you about some of
• 11:40the ways that we've interrogated
• 11:42one biomarker to look at these
• 11:44kinds of properties.
• 11:45We would expect the biomarker to
• 11:47be sensitive diagnostic status.
• 11:49I don't mean to contradict what I just said.
• 11:52I don't think that it needs to align
• 11:54perfectly with the diagnostic category,
• 11:56but if we're not seeing differences,
• 11:58at least on average,
• 11:59for example between groups
• 12:00with the condition without,
• 12:02it's really unlikely to be telling us
• 12:04something meaningful about the condition.
• 12:06Similarly, we would expect to
• 12:09see it associate with symptoms.
• 12:11We would want a biomarker
• 12:13where more extreme values,
• 12:15enough biomarker range might be associated
• 12:17with more extreme forms of symptomatology,
• 12:20maybe in different domains.
• 12:21These could be biomarkers.
• 12:23Could be specific to specific
• 12:25aspects of function.
• 12:26For example,
• 12:27we might not expect a biomarker
• 12:30to correlate with a with a gross
• 12:33measure like autism severity,
• 12:35but there might be biomarkers
• 12:37for language and biomarkers for.
• 12:39Eye contact and so on.
• 12:41In keeping with that idea,
• 12:43we would expect that a biomarker that
• 12:45we think is tracking a particular
• 12:48domain of function should be
• 12:50really specifically focused on that
• 12:52domain of function and may even be
• 12:55disassociated with other domains of
• 12:57function and let me give you an example.
• 13:00Many people with autism have
• 13:01intellectual disability,
• 13:02which is going to put a ceiling
• 13:05on their social function.
• 13:07If I had a biomarker that I thought
• 13:09was quantifying social function,
• 13:11but was actually telling me something
• 13:13about intellectual disability,
• 13:14it would do a poor job because
• 13:16it wouldn't move if I did,
• 13:18a treatment that affected social
• 13:19function but not intellectual disability.
• 13:21And so we really want to be aware
• 13:23of what a biomarker is measuring.
• 13:25And, you know,
• 13:26another way to put it would be
• 13:28we want to see that it's validly
• 13:30measuring what we think it measures,
• 13:32but it's discriminately not measuring
• 13:34things that we think aren't relevant
• 13:37to this process or system.
• 13:38In autism and pediatricians are
• 13:40accustomed to thinking this way in autism,
• 13:43we're thinking about people who
• 13:45are developing so autism can mean
• 13:47a person who is 50.
• 13:48It can mean a person who is 5 and
• 13:50so we need to think carefully about
• 13:53our biomarkers and autism in terms
• 13:56of whether a given bio mark can be
• 13:58used across the lifespan or we may
• 14:01need specific batteries of biomarkers
• 14:03for different developmental periods.
• 14:05We want biomarkers that are robust
• 14:07to variation in behavior,
• 14:09and by this I mean for some biomarkers this
• 14:12would be irrelevant for a genetic biomarker.
• 14:15It doesn't really matter.
• 14:16What's happening when that
• 14:18blood is being acquired?
• 14:19However, for the kinds of biomarkers
• 14:21that I'll talk today that are
• 14:23indices of neural function in
• 14:24an awake and behaving person,
• 14:26what's happening while
• 14:27I'm collecting that EG.
• 14:28For example,
• 14:29could actually provide more signal
• 14:30than the thing I think I'm measuring.
• 14:32If I'm trying to look at a response
• 14:34to the sound of a person's voice and
• 14:37the child is having a tantrum and
• 14:39extreme distress because of the EG net,
• 14:41well those are things that are
• 14:43going to provide more signal than
• 14:45what I think I'm measuring.
• 14:47So we want to have an awareness
• 14:48of what it is that we're trying to
• 14:50measure and what's happening in the
• 14:52context of biomarker acquisition and
• 14:53whether there can be problems in that match.
• 14:57We would like biomarkers that are
• 14:59sensitive to changes in clinical status.
• 15:01So as I said that they measure they
• 15:04might correlate with symptoms.
• 15:06Well, if symptoms change,
• 15:08do the biomarkers change?
• 15:10You know I'm going to go into practicals,
• 15:12but I want to just make a point
• 15:14that a biomarker an keeping when
• 15:16I talked about those different
• 15:18biomarkers of biomarker doesn't have
• 15:20to do all these things.
• 15:21A biomarker could be very useful
• 15:23in one context or in FDA terms.
• 15:25In one context of use.
• 15:27With some of these properties but
• 15:30lack properties that would be required
• 15:33for a different context of use.
• 15:35When we think about biomarkers
• 15:37in autism and using them for
• 15:39large scale clinical trials,
• 15:41or ideally someday using
• 15:43them for clinical practice,
• 15:44there are other considerations.
• 15:46We need biomarkers that are viable in this
• 15:49group of people who have special needs,
• 15:52people who may have sensory
• 15:54sensitivities or have difficulty
• 15:55understanding complex instructions.
• 15:57We want biomarkers that are cost
• 15:59effective that if we were to find
• 16:01something that could be useful
• 16:04in pediatricians offices.
• 16:05Could be deployed at reasonable
• 16:07cost and similar.
• 16:08We want methods that are accessible
• 16:11just to illustrate,
• 16:12let's say we had a great biomarker
• 16:15that could only be collected with a
• 16:17very specific equipment at an AT and
• 16:20NIH funded autism center of excellence.
• 16:23Well, this,
• 16:24this map depicts how many people
• 16:26it might reach in the real world.
• 16:29Conversely, if we had a biomarker that could,
• 16:33that could be.
• 16:34Deployed at any hospital,
• 16:36this is what the reach might
• 16:38be like and this is.
• 16:40This is what we seek,
• 16:42so these are the kinds of things
• 16:44that I I have in mind as we evaluate
• 16:47whether biomarkers can be useful.
• 16:49I want to tell you about,
• 16:51e.g., which some of you may know a lot about,
• 16:54but why I think it's critical for
• 16:57biomarker development and autism eegs.
• 16:59The Electroencephalogram.
• 16:59It's a measure of brain activity
• 17:02recorded directly from a person's scalp.
• 17:06When a group of neurons fire in Mass,
• 17:08the electrical signal is
• 17:10sufficiently strong that we can
• 17:12record it right through the scalp.
• 17:14Broadly speaking, two kinds of of
• 17:16activity are measured with the EG.
• 17:18What we would call resting ECG,
• 17:20which is kind of the ongoing
• 17:23oscillatory Idol of the brain.
• 17:25And then also we call event related
• 17:28potentials which are when we make
• 17:30discrete punctate things like sights
• 17:32or sounds happen over and over and
• 17:35over again and then by virtue of that
• 17:37repetition we can extract what's true
• 17:40signal from that ongoing oscillatory
• 17:42that unknown oscillatory patterns that
• 17:44can tell us something about the specific
• 17:46activity yoked to a cognitive process.
• 17:50EG is A is a great tool for autism
• 17:53because it's viable across a wide
• 17:55range of cognitive abilities and a
• 17:58wide range of developmental levels
• 18:00from literally from birth to old age.
• 18:03It's noninvasive,
• 18:04you can see here.
• 18:05This is an example of a comprehensive
• 18:07electrode net with 256 electrodes,
• 18:10and really what's what's happening
• 18:12here is that there are soft sponges
• 18:14touching a person scalp and there
• 18:17they are in salt water so that
• 18:19they create the electrical bridge.
• 18:21So the technologies that we have
• 18:23today that we use in autism are in
• 18:25advance from older technologies,
• 18:26required abrasion of the scalp
• 18:28that would be much more difficult
• 18:30and for people to tolerate.
• 18:31It's also technology that's
• 18:32pretty movement tolerant.
• 18:33These were working with
• 18:34children or people with autism.
• 18:36We need out.
• 18:37It may not expect them
• 18:38to stay perfectly still,
• 18:39and when they move we're going to lose.
• 18:42We're going to lose those trials,
• 18:43but it's not like it's going
• 18:45to ruin the entire process.
• 18:47Will still get data from the trials.
• 18:49How much a person is still,
• 18:50so it's pretty forgiving
• 18:52and accessible technology.
• 18:54It's also a very practical technology.
• 18:56It's expensive to acquire
• 18:58an EEG system like this,
• 19:00but once you have one,
• 19:02on premise is really the cost of
• 19:04acquiring data is saltwater in latex
• 19:07gloves so incredibly inexpensive,
• 19:09and it's also technology
• 19:11that's incredibly accessible.
• 19:12We think of e.g its present in.
• 19:16To my knowledge,
• 19:17every hospital in the country already,
• 19:19and it's already used at the population
• 19:22level for screening for things like
• 19:24newborn deafness and for seizures,
• 19:26and so it's a technology that should
• 19:28a biomarker that's worthy of study
• 19:31be discovered could be deployed
• 19:33at scale in a realistic fashion.
• 19:35And Lastly. EEG is a technology.
• 19:39That has been applied very affectively
• 19:41to understand social behavior and
• 19:44communication in typical development.
• 19:47And so this in autism a disorder
• 19:49that is so heterogeneous.
• 19:52It's a great advantage to have a measure
• 19:55of social function that is already
• 19:58well understood in typical develop.
• 20:00And so it gives us the opportunity to
• 20:04contrast autos people with autism from a
• 20:07set of expectations that already exist.
• 20:10This is this is an event related potential.
• 20:13Then I'm going to talk a lot about today.
• 20:16This is a face sensitive event
• 20:19event related potential called
• 20:20the N 170 so you can see highlighted
• 20:23in purple this negative dip.
• 20:25This is as I said before,
• 20:27an ERP so this is a wave
• 20:29formats extracted through signal
• 20:31processing from an ongoing e.g.
• 20:33While person is watching faces and
• 20:35what this N 170 represents is the
• 20:38very early stages of face perception.
• 20:41The phase in which it's not quite is
• 20:43this a happy face is this moms face,
• 20:45but really this is a face.
• 20:49I should say that I'll talk a lot
• 20:52about faces and looking at faces
• 20:54today and this is something that's
• 20:57of great interest in autism.
• 20:59Because faces are one of the primary ways
• 21:01that we engage in social interaction.
• 21:04Certainly probably the most effective way
• 21:06we can glean social information in infancy.
• 21:09It's an area in which people with
• 21:11autism are very well demonstrated
• 21:13to perform differently, and so,
• 21:16in a disorder that's so very
• 21:18the one commonality truly.
• 21:20Literally the one commonality in a group
• 21:22of people diagnosed with autism is,
• 21:23you know,
• 21:24that they have difficulties
• 21:25with social interaction,
• 21:26so faces and social information
• 21:27is a is a good area to go for
• 21:30in terms of trying to understand
• 21:32a very diverse group of people.
• 21:33So this these are.
• 21:34This is a waveform from the
• 21:36first study that I did.
• 21:38It was actually published during my
• 21:39first year here at Yale and this
• 21:41is what it looked like that in 170
• 21:43look like in a group of typically
• 21:46developing adolescents and adults and
• 21:47this is what it looked like in people with.
• 21:50Autism and what we found is that the
• 21:53difference that was significant wasn't
• 21:55so much an absence of signal or a
• 21:58or a complete absence of an ERP response,
• 22:01but the timing of the response
• 22:04that even at this very.
• 22:06First,
• 22:07simple stage of face processing
• 22:10people with autism showed a delay and
• 22:14inefficiency in their brain response.
• 22:17So as I go through some of the
• 22:19studies that we've done in the N 170,
• 22:22I'm going to reference that list of
• 22:24biomarker considerations that we
• 22:25talked about at the outset and so.
• 22:27We see that this N 170 measure
• 22:31latency is showing us a way in
• 22:34which at least on average,
• 22:36people with autism differ from typically
• 22:38developing people in the same study.
• 22:41We also examined persons actually
• 22:43actual ability to recognize faces using
• 22:46standard neuropsychological tests.
• 22:48And we found that despite being
• 22:52as intelligent.
• 22:53As the typically developing
• 22:54participants those with autism made
• 22:56many more enter errors nearly twice
• 22:58as many errors in face recognition.
• 23:01And then importantly,
• 23:02we saw that a person's ability
• 23:04to recognize faces.
• 23:06Was actually related to their N 170 latency,
• 23:09so we see that this N 170 latency is also
• 23:13associated with symptomatology in autism.
• 23:19The next study that we did was to try to
• 23:22understand whether this slowing was telling
• 23:24us something meaningful and at least
• 23:28relatively specific about social information,
• 23:30or whether we were measuring something
• 23:32that's just a general perceptual delay,
• 23:35and this is an important distinction,
• 23:38because we do think that many aspects of
• 23:41perception are intact in people with autism,
• 23:44and social information is
• 23:46specifically impaired,
• 23:47but that's that's one account.
• 23:49A different account, for example,
• 23:51would be that the complexity of the
• 23:53information rather than the social Ness
• 23:55of the information is what's relevant.
• 23:57So we really want to do is take things
• 23:59that are both complex and show perceptual
• 24:01differences for the social information,
• 24:03but not for the non social information.
• 24:06And this is a study that an idea that
• 24:08we got for a study by working with a
• 24:11child in our clinic and I'll show you
• 24:14a quick video to give you a sense of
• 24:17what this boy was very motivated by.
• 24:21These WS looks like they are on top of
• 24:25each other. We sure do you a
• 24:30picture. Yeah. PS. That's.
• 24:39This level is not underwater. He's got
• 24:45a mask and a snorkel. Anyway, just like Jay.
• 24:52You're right, it does look like a J.
• 25:00OK, so so clearly this is a boy who
• 25:02is very interested in letters and
• 25:05I'm going to talk more about that.
• 25:07I will make a comment as an aside that
• 25:10I started out by saying that we lack
• 25:13biological tools to assess an autism
• 25:15and this is an example of it that
• 25:18play session that you just witnessed.
• 25:20That's a part of the gold standard
• 25:23diagnostic tool for autism and so just to.
• 25:26An illustration of how behaviorally based
• 25:28our current ways of understanding autism,
• 25:30but to get back to this boy
• 25:33in this experiment,
• 25:34this gave us the idea to contrast neural
• 25:37processing for social information in the
• 25:39form of faces with perceptual processing
• 25:42of non social information for letters.
• 25:44It actually turns out that when
• 25:46one learns to read an alphabet,
• 25:48any alphabet you actually develop an
• 25:51N 170 to letters of the alphabet that
• 25:54tends to be lateralized differently.
• 25:56In the faces faces on the right,
• 25:58letters on the left,
• 25:59and so we compared faces to houses both
• 26:01visually complex but one social one not
• 26:04and then letters to a made up alphabet.
• 26:06And So what we would expect is that if
• 26:08it's a problem with social perception,
• 26:11we would see differences here in people
• 26:13with autism but intact performance here.
• 26:15But if it's general perceptual problem,
• 26:16we would see differences in both places.
• 26:20In terms of and this also we also
• 26:21did in this study was took a look
• 26:24at development and so this was a
• 26:26younger cohort than those that we
• 26:28studied the Edmond 17 before to see
• 26:30whether the differences that we're
• 26:32seeing are still consistent.
• 26:33And in terms of the social,
• 26:36the social aspects they were,
• 26:37we saw what we'd seen before that people with
• 26:40autism had worse face recognition scores.
• 26:43We saw other than 170 is
• 26:45slower in people with autism,
• 26:47so we're finding that this end 170
• 26:49latency delay that we had discovered in
• 26:52adolescents and adults is actually holding
• 26:54up even in much younger people in children,
• 26:57and actually studies that have
• 26:59happened subsequently have shown
• 27:01in children as young as three.
• 27:04When we look at the non social information,
• 27:06the letter processing in terms
• 27:08of the behavior,
• 27:09we found that there weren't
• 27:11behavioral differences.
• 27:11People with autism showed
• 27:13normative reading scores,
• 27:14and then if we if we broke things
• 27:16down into even phoning decoding,
• 27:18they were again comparable to their
• 27:20typically developing counterparts.
• 27:21And when we looked at their brain
• 27:24responses so you can see in purple
• 27:26obrein response to a letter and in green
• 27:29of brain response to a pseudo letter,
• 27:31we're seeing that in both groups this.
• 27:34Larger amplitude response because
• 27:35it's a negative response.
• 27:37It's going down is showing specialization
• 27:39in both groups were not seeing differences
• 27:41in Specialization in terms of the power,
• 27:43and we're also not seeing differences
• 27:46in Specialization in terms
• 27:47of the timing of N 170.
• 27:49The only difference that we saw between
• 27:51the groups was at the group with
• 27:53autism showed atypical lateralization.
• 27:55They they were relying more on the face
• 27:57related regions in the right hemisphere
• 27:59than the typically developing children.
• 28:01So this was important because it showed us.
• 28:04That the differences that we were seeing
• 28:07in the M170 would not just a reflection
• 28:10of a generic perceptual problem,
• 28:12but are really telling us something
• 28:15specific about perception of faces.
• 28:19The next question we had was whether the
• 28:22way that people with autism engage with
• 28:25faces could be accounting for our results.
• 28:28So people with autism tend to look more.
• 28:32Tend to look less to the eyes of the
• 28:34face and often look more to the mouth,
• 28:37or as most typically people
• 28:39really gravitate towards the eyes,
• 28:40and this is relevant because where
• 28:42you look on a face can actually
• 28:44modulate the speed of your N 170
• 28:46because it's faster to the eyes.
• 28:48And so if we did an experiment and
• 28:51people with autism are not looking
• 28:52at the faces in the same way,
• 28:55maybe they're just looking at the mouse.
• 28:57Well, that could elicit longer
• 28:58and 170 latencies,
• 28:59but it's not really telling us anything
• 29:02about the way that people with autism.
• 29:04The brains are responding to the face.
• 29:06Is it telling us a difference in the
• 29:09way their eyes are taking in the
• 29:11faces and so we measured this we.
• 29:13We used a crosshair to control the
• 29:16persons visual attention and just
• 29:18basically forced people to look either
• 29:20to the eyes to the nose or to the Mount,
• 29:23and we wanted to see is if we forced people
• 29:26with autism to look to the eyes of the face.
• 29:30Would these inefficiencies
• 29:31in phase processing go away?
• 29:35And they did. What we found is that if you
• 29:38can see here in blue is most important.
• 29:42The eyes that typically developing children
• 29:45shown in darker blue and the children
• 29:47autism shown in lighter blue is that rather
• 29:50than normalizing bringing activity in ASD,
• 29:53drawing attention to the eyes make these
• 29:56differences more pronounced that typically
• 29:58developing people show this dream.
• 30:00Attic enhancement in and 170 latency
• 30:03rapping rap quickening event 170 latency
• 30:05to faces but people with autism don't,
• 30:07so we infer the end 170 delays that
• 30:10we've seen across the studies that
• 30:12we've described so far are not just
• 30:15an artifact of the way a person with
• 30:18autism looks at the stimuli on screen,
• 30:21but are actually telling us something
• 30:24about the neural systems that we think
• 30:27relate to autism symptomatology.
• 30:29And then the last kind of scientific aspect
• 30:32of the end 170 that we've interrogated.
• 30:34And this is these are very
• 30:36preliminary data there under review.
• 30:38Now it's a very small sample,
• 30:40but it's also very exciting data to
• 30:43us in that no data of this nature
• 30:46exist yet in autism.
• 30:47And really, this is data looking
• 30:50pre post at changes in the end,
• 30:52170 in parallel to improvements in behavior.
• 30:54And so this is a collaboration with
• 30:57Pamela Ventola and the faculty.
• 30:59The child Study Center.
• 31:00Who's in.
• 31:01Expert in a behavioral treatment for
• 31:03autism called pivotal response treatment.
• 31:05It's it's a well studied a well
• 31:08validated treatment that works.
• 31:10It focuses on social skills over
• 31:12the course of 14 weeks and we would
• 31:15theorize that it should target the
• 31:18the systems that the N 170 measures
• 31:21face perception and social motivation.
• 31:24As Pam did this treatment,
• 31:26kids got better behaviourally and we
• 31:28looked at how their ERP's changed,
• 31:30and So what you're seeing here on
• 31:32the left is a pre measure on the
• 31:35right is a post measure.
• 31:36After these 14 weeks of treatment and
• 31:39you can see that each line represents
• 31:41an individual child an the dotted
• 31:43line represents the average and so
• 31:45again this is only seven children.
• 31:47But what we're seeing is that on
• 31:50average the end 170 is getting faster.
• 31:52This latency delay is being reduced.
• 31:54As kids get better and it's really
• 31:56true for the individuals borrowing
• 31:58participant #3 who we don't.
• 32:01We don't have any nothing in the phenotype
• 32:03makes clear why that would be different.
• 32:06So again, suggestive evidence that N 170
• 32:10latency may also index clinical status.
• 32:14So let's review.
• 32:15We've talked, we've done all these
• 32:18different studies at the end 170.
• 32:20We've seen that it's sensitive
• 32:22to diagnostic status.
• 32:23It's associated with symptoms.
• 32:25It is functionally specific.
• 32:27It's applicable across a
• 32:28wide developmental range.
• 32:29It's robust to variations in behavior
• 32:32during biomarker acquisition.
• 32:33It's sensitive to changing clinical status,
• 32:35and if you remember a
• 32:38practical considerations,
• 32:38it is viable in this population.
• 32:41It is cheap, and it is iaccessible.
• 32:44So lots of strong evidence.
• 32:46Why am I saying that we have
• 32:49no biomarkers for autism?
• 32:51Well, first I want to make clear that the the
• 32:55the problems that I see with the N 170 and
• 32:58many of the benefits that I see the N 170.
• 33:01Although it is one of the most well
• 33:04studied markers at this point,
• 33:05are true of many biomarkers and autism.
• 33:08There are many biomarkers
• 33:09collected via neuroimaging.
• 33:10The eye tracking that show promising
• 33:12attributes in many of these domains,
• 33:14but a truth for all of them is that there's
• 33:16inconsistent reproducibility at you.
• 33:18See now at the bottom of
• 33:20the slide every study.
• 33:22That has followed up on our initial findings
• 33:24of the N 170 since we published it in 2004,
• 33:27and as you can see, there are many.
• 33:30There's even a great meta analysis published
• 33:33in 2017 that shows across all of these.
• 33:36You know, on average,
• 33:37people with autism are slower
• 33:39in their end 170,
• 33:40and this specific to faces but not generic.
• 33:42But there are studies among these
• 33:44that fail to find those differences,
• 33:46and we don't know why.
• 33:49It might just be the truth in a
• 33:51disorder is heterogeneous autism.
• 33:53Maybe there are many ways
• 33:55to arrive at autism,
• 33:56and some of those pathways don't
• 33:58impact the face perceptual system.
• 33:59So maybe some of these studies with
• 34:01null findings are telling us that they
• 34:04measured an actual cohort of children
• 34:06with autism whose face processing is fine.
• 34:11Maybe another problem though is
• 34:13that many of these studies are small
• 34:16and so some of these non results
• 34:19could reflect underpowered studies.
• 34:22And Lastly, we don't know how differences
• 34:24in methodology among many of these studies.
• 34:27For example what e.g system you use,
• 34:29whether you show neutral smiling
• 34:31black and white color faces,
• 34:32how these things impact in N 170.
• 34:35So there's lots of mythological
• 34:36noise in these datasets as well.
• 34:38That and we don't know.
• 34:40We in the field, not just autism.
• 34:42In cognitive neuroscience we
• 34:44don't know how many of those
• 34:46differences would affect in any 170.
• 34:48We also know very little about the
• 34:51reliability of a measure like this,
• 34:54where many most biomarkers
• 34:55for autism within a person.
• 34:57Overtime,
• 34:58we don't know whether the act
• 35:00of taking repeated measurements
• 35:02changes the values of the biomarker,
• 35:04and these are things that are critical
• 35:07to know for use in clinical trials.
• 35:10It's it's meaningless to take
• 35:13repeated measurements of a biomarker
• 35:15that isn't stable overtime.
• 35:17We we lack in normative reference and
• 35:19by that I said earlier that we have
• 35:22an understanding of ERP's and Social
• 35:24Development from typical cognitive
• 35:26neuroscience, and that's true.
• 35:28But I mean,
• 35:29in pediatricians are great group
• 35:31to make this point too,
• 35:32because when a child comes into your
• 35:35office you can look at a growth chart
• 35:38and know exactly where that child
• 35:40falls in terms of their percentiles.
• 35:42And we don't have anything
• 35:44like that when we say in 170 is
• 35:47delayed in a person with autism.
• 35:49All we know that it is delayed relative to
• 35:51the control group in that particular study.
• 35:54There's no kind of atlases
• 35:56for making inference.
• 35:57And So what we need all of these things,
• 36:00I think,
• 36:01can be addressed,
• 36:02but we need are more rigorous
• 36:03biomarker studies really designed
• 36:05to collect the kind of information
• 36:07that's needed for FDA qualification.
• 36:08What would these studies look like?
• 36:10They would be testing well,
• 36:12evidenced biomarkers.
• 36:13This is this sounds intuitive,
• 36:14but it's hard for those of
• 36:16you who are grant writers.
• 36:18You know it's going to be a lot easier
• 36:21to fund the first study of something
• 36:23that no one's ever heard of than
• 36:26it is to fund the 3040 first study.
• 36:28Of the N 170,
• 36:30but that's what's needed,
• 36:31right?
• 36:32Is deep interrogation of the most
• 36:34promising and well studied biomarkers.
• 36:36We want cohorts that are very
• 36:38well phenotyped with existing
• 36:39clinical measures so that we could
• 36:42understand the relationship between
• 36:44the biomarker and symptomatology,
• 36:45and so that we can tease
• 36:48apart individual differences.
• 36:50We want large samples,
• 36:51including large samples of typically
• 36:54developing control children so
• 36:55that we can understand kind
• 36:57of a normative reference.
• 36:59We want a design that lets us take
• 37:02repeated measurements overtime
• 37:03so that and I don't mean years.
• 37:05Often in psychology we think of
• 37:07lunch tunell meaning following
• 37:09a child through adulthood,
• 37:10I mean following a child through the
• 37:12length of a typical clinical trial.
• 37:14So six months.
• 37:17We want studies that do away
• 37:19with as much as
• 37:20possible the kinds of methodological
• 37:23inconsistencies and noise that
• 37:25plagued the literature thus far.
• 37:28We want practical assays.
• 37:31And these are all the all the factors
• 37:34that went into an RFA that was published.
• 37:37Now about six years ago to fund a
• 37:40consortium to develop biomarkers and autism.
• 37:43And this is the work that we've done
• 37:45under the name of the optimum biomarkers
• 37:48consortium for clinical trials.
• 37:50So this is a multi site naturalistic study
• 37:53despite being called for clinical trials.
• 37:55It's not a clinical trial.
• 37:57It's based here at Yale, largely NYC.
• 38:00I I won't go through all
• 38:02the different places, but.
• 38:04It involves five major autism
• 38:06research centers around the country.
• 38:09Has a data coordinating core that's also
• 38:11based at Yale and then a distributed
• 38:14data acquisition analysis core.
• 38:15Really pulling in key expertise at at
• 38:18different sites around the country.
• 38:20We saw a large group of children
• 38:23280 with autism,
• 38:24119 with typical development between
• 38:25the ages of 6 and 11 with a wide
• 38:29range of IQ going well into the
• 38:31intellectually disabled range.
• 38:33We focused on practical assays
• 38:35like EEG and eye tracking.
• 38:37And we use the longitudinal design,
• 38:39A baseline a six week,
• 38:41and a 24 week to let us look at Test
• 38:43retest reliability or I we shy away
• 38:45from the term test retest 'cause not
• 38:48instantaneous but short term stability
• 38:50and then change over six months
• 38:52when children will have grown when
• 38:54they may have received treatments.
• 38:55And then we also acquire genetic information.
• 38:57All these children.
• 39:00There are a few other unique
• 39:02aspects of this study.
• 39:04It's a.
• 39:04It's a cooperative agreement,
• 39:06au 19 mechanism, and so really this
• 39:09study is run not just by Apiai,
• 39:11but by a steering committee
• 39:13involving the government,
• 39:14government, scientists,
• 39:15academia, and actually scientists
• 39:17from industry as well.
• 39:18We applied truly an unprecedented,
• 39:20at least in autism, neuroscience,
• 39:22research, level of rigor.
• 39:24We administer the study,
• 39:25even though it's not a clinical
• 39:27trial according to good clinical
• 39:29practice standards methodologically.
• 39:31We made sure that from the from the
• 39:33software used to the cables used to
• 39:36connect the monitor to the computer.
• 39:38That hardware was identical.
• 39:40Software is identical,
• 39:41everything was processed centrally.
• 39:42So tremendous amount of detail to
• 39:45biomarker validity and then statistically,
• 39:47really we collected so much data data.
• 39:49But because we wanted to bring this
• 39:52to the FDA we pre specified our
• 39:54hypothesis in very, very specific ways.
• 39:57So really tons and tons and
• 39:59tons of information.
• 40:00Is being distilled to a T test
• 40:03with directional hypothesis.
• 40:04And then Lastly,
• 40:06we harmonized our study with a
• 40:08consortium in Europe called the
• 40:10EU aims so that we would have
• 40:12the opportunity to both validate
• 40:14results and compare across samples.
• 40:18These are the biomarker essays that we used.
• 40:21I'm only going to talk about 2:00 today.
• 40:23One you could have guessed the
• 40:25ERP's to faces which derives the
• 40:28N 170 that we've talked about and
• 40:30then a composite that we created
• 40:32across three different bio mark.
• 40:34Three different.
• 40:35I different eye tracking assays,
• 40:37a composite telling us how much a person,
• 40:40basically how much child with autism looks
• 40:42too faces or people when they're on screen
• 40:45and we call that the Ocular Motor Index.
• 40:48Of gays to human faces.
• 40:51This is again I won't enumerate all of these,
• 40:54but these are.
• 40:55This is this is where we live.
• 40:58This is the current status quo.
• 41:00The clinical measures that we
• 41:02use in our research studies.
• 41:03We missed all of these both to give
• 41:06us a kind of a frame reference,
• 41:09but then also allow us to examine
• 41:12correlations with the phenotype.
• 41:14In terms of M170 Latency,
• 41:16I think you all at this point could
• 41:19could into it where we predicted we
• 41:21predicted that we would see a slower
• 41:24an 170 in this sample and we did
• 41:26things that we learned where that we
• 41:28can collect this information very
• 41:30reliable reliably in young children
• 41:32you can see we got valid signal
• 41:34a 97% of a typically developing
• 41:36in 3/4 the children with autism.
• 41:38If people are interested we can
• 41:40talk on the question and answers.
• 41:42That is that is a low number.
• 41:45Due to an unnecessarily difficult
• 41:47experimental paradigm,
• 41:48we could get this data with a
• 41:50quicker paradigm that would likely
• 41:52increase increase acquisition ASD.
• 41:54We also learned that it's fairly stable,
• 41:57so six weeks stability in the typically
• 42:00developing children is .75 in autism.
• 42:03It's .66,
• 42:03and then we also saw again that
• 42:06this relates to the phenotype.
• 42:09In some ways that we anticipate
• 42:11in some ways that are novel.
• 42:14We see that again the end 170 is
• 42:17related to your face memory skill.
• 42:19We also saw that it's related to
• 42:21an aspect of adaptive functioning
• 42:23acquired from a parent interview
• 42:25called the Vineland adaptive behavior
• 42:27scales abbreviated vabs here,
• 42:29and it's a scale associated with
• 42:31your kind of drive to interact
• 42:33with other children and to play.
• 42:36I won't get into the P 100,
• 42:38but there's other aspects of the
• 42:40waveform that we can examine and
• 42:43we when we expand beyond the 170,
• 42:45we also see that.
• 42:47Other aspects of the face perceptual
• 42:49system correlates with the phenotype.
• 42:52In terms of the Ocular Motor Index,
• 42:54similarly,
• 42:55we predicted that people with autism
• 42:57would look less to faces and that was true.
• 43:01We saw that we get almost perfect
• 43:03acquisition with eye tracking.
• 43:05It's a very,
• 43:06very tolerable procedure for
• 43:08people with autism.
• 43:09We saw that it's highly stable
• 43:12across six weeks.
• 43:13.8 three about in both groups,
• 43:15and again we saw that it was
• 43:18related to the autism phenotype.
• 43:22So with this information we took a
• 43:24step and we submitted letters of
• 43:26intent to for both biomarkers to the
• 43:29FDA's biomarker qualification program.
• 43:31Both were accepted one in May 19,
• 43:342019, one in March 2020.
• 43:36This is let me give you a little bit
• 43:39of context about what this means,
• 43:42so this does not mean that they're
• 43:44qualified biomarkers by any stretch.
• 43:47Really,
• 43:47it means the FDA recognizes
• 43:49the value and the
• 43:51rigor of the data that we provided.
• 43:54And they think that this is
• 43:56worth worth studying and so.
• 43:58What this means now is that we are
• 44:00actually preparing what's called
• 44:02a biomarker qualification plan,
• 44:04which is a detailed set of studies
• 44:06and analysis that will provide the
• 44:08additional information that the
• 44:10FDA would require to qualify them,
• 44:12and should we be successful in that,
• 44:15the next step is a biomarker
• 44:17qualification package.
• 44:18Or we basically submit the
• 44:20results of those studies,
• 44:21and the FDA decides whether this
• 44:24can be a qualified biomarker.
• 44:27While we're at the first step of the journey,
• 44:30I do want to highlight that it's an
• 44:33incredible milestone for our field.
• 44:35Really, for the field of psychiatry,
• 44:37and that these are the first 2 letters
• 44:40of intent submitted to the FDA or
• 44:42were accepted to the FDA for autism,
• 44:45or for any psychiatric disease.
• 44:47And so we have a long,
• 44:49long way to go before we have
• 44:52qualified biomarkers.
• 44:53At the same time,
• 44:54we're also in completely uncharted
• 44:56and very exciting territory.
• 44:58Working with the FDA to understand
• 45:00how to qualify qualified biomarkers
• 45:02for these kinds of conditions,
• 45:04which are really different than things
• 45:06like a tumor or or Alzheimer's,
• 45:09which are are areas in which
• 45:11this biomarker kind of research
• 45:14has been much further along.
• 45:17In terms of how we pitched this to
• 45:19the FDA is as a diagnostic biomarker.
• 45:22Clu stands for context of use,
• 45:24and so really this biomarker qualification
• 45:26process is specific to a context of use,
• 45:29but when we when we describe this
• 45:31to the FDA's diagnostic biomarker,
• 45:33if we were to come back and say no,
• 45:36no,
• 45:36we think this is going to be
• 45:38useful as a prognostic biomarker.
• 45:40It would be a whole new letter of intent,
• 45:43so these these processes are
• 45:45catalyzed the idea.
• 45:46Is not that it's diagnostic biomarker
• 45:48telling us who has autism or not,
• 45:51but this is these are data from the
• 45:53MI and you can see the distribution
• 45:55of typical of people with autism.
• 45:58Lower is less attention to faces
• 45:59and you can see the distribution of
• 46:02people who don't have autism and you
• 46:04can see this portion of the group
• 46:07of people with autism highlighted
• 46:08in purple and that's basically what
• 46:11we're saying is that we think that
• 46:13this could be a meaningful stratum
• 46:15within the category of autism.
• 46:17Indicating a more homogeneous
• 46:19etiology or similar phenotype,
• 46:21neural phenotype and that by
• 46:24enriching clinical trials with people
• 46:27in this part of the distribution
• 46:30we would reduce heterogeneity and
• 46:33have more powerful trials.
• 46:35That's the way that we're thinking about it.
• 46:38We received different grants,
• 46:40cooperative agreements to work
• 46:42with the FDA on both of these
• 46:45biomarker qualification plans.
• 46:46It's very much a collaborative
• 46:48process and the kinds of things that
• 46:51we're wrangling with now are really,
• 46:54how do,
• 46:54how do we determine cut points
• 46:56in a continuous distribution?
• 46:59How do we verify that these strata
• 47:02are meaningful in some way?
• 47:04We really need some kind of
• 47:06external reference to say
• 47:08that this stratification.
• 47:10Is something useful and it's hard
• 47:11because we don't have tremendous
• 47:13faith in our clinical measures alone,
• 47:15and so it's really challenged.
• 47:17And then Lastly,
• 47:18we're we're working with them to
• 47:20think about how to demonstrate
• 47:22that these differences that we
• 47:24see so reliably are not just
• 47:26reflective in some way of the
• 47:28particulars of our acquisition setup.
• 47:32As Linda mentioned, we're very,
• 47:34very excited that we were renewed
• 47:37for another five years in July.
• 47:39The work that we're undertaking is going
• 47:42to involve three separate studies.
• 47:45The first study is a confirmation
• 47:47study where we're basically going to
• 47:50replicate the results of our first study,
• 47:52again, 400 children.
• 47:54We're going to actually increase the
• 47:57proportion of typically developing children
• 47:59so that we can get a stronger kind of.
• 48:02External reference in terms
• 48:04of some of the measures,
• 48:06same age, same logical design,
• 48:08same biomarker batteries accepting
• 48:10biological motion which was not
• 48:13effective in our batteries and
• 48:15we're going to eliminate just
• 48:17to reduce participant burden.
• 48:19We're also going to do a follow
• 48:21up study in which we bring back
• 48:23the original cohort from the first
• 48:26phase of the EBCT and evaluate them
• 48:28at what will be 2 1/2 four years
• 48:31after their initial enrollment,
• 48:32and this will get us give us a chance
• 48:35to look at a few different things.
• 48:38So how stable these biomarkers
• 48:39are over the longer term?
• 48:41Conversely, if children aren't stable,
• 48:43how well these biomarkers track
• 48:44with changes in clinical status?
• 48:46And Lastly, a sense of Lanja Tude Inal.
• 48:49Addictive value so we will have
• 48:51a biomarker value.
• 48:52You know at one point one time point
• 48:55and will have it for years later
• 48:57in some kids and to what degree are
• 49:00initial values predictive of either
• 49:02future biomarker values or or future
• 49:05clinical status or phenotypic values?
• 49:08The third study is a feasibility
• 49:10study and it will involve basically
• 49:12taking a much smaller cohort.
• 49:1425 children autism 25 typically
• 49:16developing children and seeing if
• 49:18we can successfully deploy this
• 49:19battery in three to five year old
• 49:22children just at one time.
• 49:23Really just seeing if it works
• 49:25and that's where we stand.
• 49:27We're really excited that we have the
• 49:30opportunity to carry this work forward.
• 49:33I wanted to close,
• 49:35you know if the ABC T so it really
• 49:38is true and giving this talk
• 49:40in preparing this talk,
• 49:43I reflected that the research
• 49:45that I've described to you today
• 49:47really reflects the arc of my
• 49:50scientific career in its entire T,
• 49:52and it's really about the Abcte,
• 49:55at least is about scope
• 49:57rather than innovation.
• 49:58That I would argue actually this.
• 50:00Scope is the innovation for that study,
• 50:03but we are trying to innovate and
• 50:05one of the ways that we're excited is
• 50:08whether we can use these biomarkers
• 50:10to guide treatments and so one way
• 50:12would be to think about direct
• 50:15brain stimulation.
• 50:15So when we think about the behavioral
• 50:18treatments that are really pretty
• 50:20much all that exist for autism,
• 50:22they target social function and we
• 50:23know from some imaging studies that
• 50:25their reflected in altered activity
• 50:27in the posterior superior temporal sulcus,
• 50:30a brain region.
• 50:31Implicated in autism and also
• 50:33implicated in social perception.
• 50:36And So what we're trying to
• 50:38understand now is if whether it,
• 50:40whether we can use transcranial magnetic
• 50:42stimulation to directly activate the
• 50:44STS and then a longer term goal,
• 50:46would be to see if it improves function in
• 50:49people with autism in a therapeutic way.
• 50:52But a short term goal would be to
• 50:55leverage these biomarkers that we have
• 50:57and to see whether stimulating the.
• 51:00Attenuates the 10170 latency delay
• 51:02or whether it improves attention
• 51:04to faces and eyes in eye tracking,
• 51:06and we've been in the very early
• 51:09stages of this line of work.
• 51:11We're we're collecting pilot data through
• 51:14the support of the slip foundation
• 51:16and what we found so far is that.
• 51:19It seems to have those effects,
• 51:21even people who don't have autism.
• 51:23So when we we test these procedures
• 51:25on people with typical development,
• 51:28we see that there anyone 70s get
• 51:30faster after stimulation of the tests,
• 51:32and we see that even in people
• 51:35who have normal I looking,
• 51:37they tend to look more to the
• 51:39eyes after after treatment,
• 51:41and so I shouldn't stream after
• 51:43stimulation and so we do think
• 51:45that this is a very promising Ave.
• 51:48There is outstanding resources
• 51:49in neuromodulation at Yale and
• 51:51we're looking forward to.
• 51:53Carrying this work forward
• 51:54and I'm going to end there.
• 51:56I want to acknowledge a few groups.
• 51:59Most importantly the children,
• 52:00the families, the adults,
• 52:01adolescents who come in and
• 52:03spend their time wearing,
• 52:05e.g.
• 52:05Nets to help us learn.
• 52:07We are extremely grateful
• 52:08for that partnership.
• 52:09This is also a partnership I've
• 52:11really talked about mostly about
• 52:13the laboratory side of things,
• 52:15but this is inextricably tide to
• 52:17the clinical work that we do,
• 52:19and so I want to acknowledge
• 52:21the clinicians in the other
• 52:23element Disabilities Clinic.
• 52:25I want to acknowledge the broad group
• 52:27of investigators involved in the autism
• 52:29biomarkers consortium for clinical trials.
• 52:31This is a very small subset of them,
• 52:34only indicating leadership in the consortium,
• 52:36so it's a It's a big group who's worked
• 52:39very hard to conduct a very ambitious study,
• 52:42and then Lastly,
• 52:43the folks right here at home.
• 52:45The people in the lab who are
• 52:48doing this work day today,
• 52:49so I will stop there and I'm happy
• 52:52to entertain any questions in
• 52:54the time that we have remaining.
• 52:57Great, thank you so much Jamie.
• 53:00As a pulmonologist I have to say
• 53:02that your talk with spectacular an
• 53:04I just gained so much insight and I
• 53:07know we've already got a couple of
• 53:10questions we have time for just a few,
• 53:13but I'd like to invite OTA
• 53:16fennec for the first one.
• 53:18Thank you so thank you for
• 53:20this really terrific talk.
• 53:21This has been very really very
• 53:22interesting both on the professional
• 53:24and on the personal level,
• 53:25and I'm hopeful that one day
• 53:27there's going to be a biomarker
• 53:28that we can use in primary care.
• 53:30I'm wondering,
• 53:31given what you've shown and
• 53:32what you know about facial
• 53:33recognition and attention to eyes,
• 53:35what do you think the effects of
• 53:37facial masking related to kovid will
• 53:38be on the development of children?
• 53:40Either the typically developing
• 53:41ones or children with autism?
• 53:42Because the kids are now seeing a lot
• 53:45of eyes and not a lot of the rest
• 53:47much of the time outside the home.
• 53:49I'm thinking what are the effects of that?
• 53:52It's a great question in it.
• 53:55I I really I could make up lots of answers.
• 53:58I don't know. I can say this.
• 54:01I can say that the face processing system
• 54:04is an exceptionally plastic system.
• 54:07We you know. Infants process non
• 54:12human primate faces very similarly.
• 54:16To human faces, and as we grow and
• 54:18get better at processing faces,
• 54:20we see that we no longer treat monkey
• 54:23faces like human faces and what
• 54:24that's telling us is that a system
• 54:27starts out with a really general,
• 54:29broad approach to perceiving something
• 54:31and then gets better at the things
• 54:33that are most important and gets and
• 54:35loses some of the skill in the aspects
• 54:37that are that are least important.
• 54:39And that's the way I think about the
• 54:42world that we're living in today.
• 54:44I don't? I don't, I don't know.
• 54:46But my sense is that you know most
• 54:49of the information I have, this.
• 54:51I'm glad we're wearing the masks over
• 54:53our mounds and not over our eyes,
• 54:55because really,
• 54:56there's a tremendous amount of
• 54:58information even affect if an
• 55:00emotional information that can
• 55:01be communicated through the eyes.
• 55:02And my suspicion is that maybe maybe
• 55:05old timers like me are going to
• 55:07have a much harder time inferring
• 55:09facial expressions,
• 55:10but my guess is that the young and
• 55:12plastic brains are going to become
• 55:14just that much better at inferring.
• 55:16A lot of this information.
• 55:18From the eyes I will say,
• 55:20I think that the aside from affect,
• 55:23I think a bigger challenge maybe.
• 55:26Speech processing and that many people
• 55:29with autism tend looking to the mouth,
• 55:32and autism tends to associate with.
• 55:35Stronger language,
• 55:36and so there has been the idea that
• 55:38people with autism look to the mouth
• 55:41to help them understand things and so
• 55:43that might be more of a challenge.
• 55:45You know, there's I don't think
• 55:47there's a data to support it yet,
• 55:49but those kinds of things could be
• 55:51addressed with transparent math masks,
• 55:53but it's a really good question where
• 55:55certainly curious about it in our studies,
• 55:57and that this is a study built,
• 56:00largely run replication,
• 56:01and we can't do the things
• 56:03that we did the first time.
• 56:05We're going to do things differently,
• 56:06and so.
• 56:07Hopefully we'll understand
• 56:08more in the next few years.
• 56:10Yeah, it's going to be interesting to
• 56:12see what the differences are. That's
• 56:13really cool. Thank you.
• 56:16OK, thank you so much and Jamie
• 56:18just quite well. Two questions.
• 56:20One is on the specific.
• 56:23Sorry specificity of the end 170
• 56:26latency delay or there any other causes.
• 56:29Thinking about it in future clinical utility.
• 56:34It's a great question.
• 56:35There's very there's most studies.
• 56:38Compare children and adults with
• 56:40autism to typical development.
• 56:41We've done a study comparing
• 56:43adults with autism to adults with
• 56:45schizophrenia and schizophrenia also
• 56:47disrupts the face processing system,
• 56:50and we do see differences in their
• 56:53brain response that look different.
• 56:55They aren't the same delays.
• 56:57Similarly, social anxiety is
• 56:59impacts the face processing system.
• 57:01You see different differences
• 57:03in Edwin 70 amplitude.
• 57:05People with anxiety that the
• 57:07delays in latency seem to be
• 57:09relatively specific to autism, but.
• 57:13Based on, you know,
• 57:15compared to the kind of data that
• 57:17we have in the ABC, tiebout large,
• 57:20very rigorously collected samples,
• 57:21there's there's a much weaker,
• 57:23much weaker foundation for inference,
• 57:25but it's certainly something
• 57:26to be aware of it,
• 57:28so you can have autism and have other things,
• 57:31and so you can have autism.
• 57:34You can have anxiety,
• 57:35you can have autism,
• 57:36you can have ADHD,
• 57:38so part of the process of of
• 57:40getting these biomarkers ready for
• 57:42prime time will be understanding
• 57:44just those things OK?
• 57:46If we do establish it is something
• 57:48that is different in autism.
• 57:49Is it uniquely different autism
• 57:50or to what degree is it telling us
• 57:52just about a system is disrupted?
• 57:55OK, thank you and for our last
• 57:58question I'll hopefully run
• 58:00if you're on Doctor Angoff.
• 58:02Hi. Jamie running off height.
• 58:07So I was wondering,
• 58:08I'm just thinking in terms of
• 58:09whether there's any thought of
• 58:11looking at whether.
• 58:12The kind of intensive ABA therapy
• 58:16that that we use so much in kids
• 58:20with autism would would be.
• 58:24Something you could look at the
• 58:28The N 170 latency.
• 58:31Yeah, very much so.
• 58:33In fact, the preliminary data that
• 58:35I showed from part pivotal response
• 58:37treatment is a form of applied
• 58:39behavior analysis or ABA therapy.
• 58:41It's just a form of design to
• 58:44be more naturalistic than more
• 58:45traditional discrete trial approaches.
• 58:47It's interesting. It's certainly.
• 58:49I think it's certainly worth studying
• 58:52the questions for me would be.
• 58:53What are the treatments address
• 58:55and what is the end 170 index?
• 58:58And you know a lot of not all.
• 59:01With eyes appear T in comparison with
• 59:0470s party is really geared towards
• 59:06driving childrenswear social information,
• 59:09whereas ABA can have a much broader
• 59:11range of targets and something
• 59:13including some kind of purely academic.
• 59:16And it's a It's a good question,
• 59:19question,
• 59:20question being would treatment that
• 59:22improves autism symptomatology in
• 59:24a Broadway but doesn't specifically
• 59:27target social perception change
• 59:28the N 170 and there you know,
• 59:31there's data from party.
• 59:32They not the data they showed
• 59:34today there only that have.
• 59:36Pre and Post with ERP there's pre and
• 59:38Post with F MRI that shows increased
• 59:41activity in the TS and there's a
• 59:43study with ERP that didn't have
• 59:45pre data but showed that the after
• 59:48a different behavioral treatment
• 59:49called early start Denver model
• 59:51which I would still categorize
• 59:53in ABA kind of approach in this
• 59:55also kind of socially oriented.
• 59:58They did not.
• 59:59There weren't significant difference.
• 01:00:01This is between groups after treatment,
• 01:00:02however, because there wasn't predate,
• 01:00:04it's hard to know exactly what that means.
• 01:00:06But yeah, it's it's an area that what you
• 01:00:09suggested exactly needs to be studied.
• 01:00:11The study the study,
• 01:00:12just haven't all been done yet,
• 01:00:14it's just.
• 01:00:14It's hard.
• 01:00:15It's hard to do a lot of these studies.
• 01:00:20Great, well thank you so much Jamie.
• 01:00:22We really really enjoyed having you
• 01:00:24to pediatric grand rounds and look
• 01:00:26forward to future collaborations.
• 01:00:28Thank you. Thank you so much. Really
• 01:00:30my pleasure. Thank you for having me.