YC-SCAN2 May 2026 Webinar

Transcript

• 00:04Okay.
• 00:05Welcome, everyone. It's, my great
• 00:08pleasure to introduce today,
• 00:11two speakers.
• 00:12First,
• 00:14doctor Rachel Rabien. She's an
• 00:16assistant professor in the department
• 00:17of psychiatry
• 00:19at McGill University
• 00:20and researcher at the Douglas
• 00:22Research Institute where I work.
• 00:25And then, we'll be hearing
• 00:27about doctor Renato
• 00:29Polimanti,
• 00:30who's an associate professor of
• 00:31psychiatry
• 00:32in the biomedical informatics and
• 00:34data science and chronic disease
• 00:36epidemiology at Yale School of
• 00:38Medicine.
• 00:39So,
• 00:40first, I'll introduce,
• 00:42Rachel.
• 00:43She leads the Addiction Imaging
• 00:45and Mental Health Laboratory
• 00:46where her work focuses on
• 00:48understanding the clinical,
• 00:50cognitive, and neurobiological
• 00:52mechanisms underlying substance use and
• 00:55mental health disorders.
• 00:57Her research integrates advanced neuroimaging
• 00:59techniques,
• 01:00clinical assessments,
• 01:02and translational neuroscience
• 01:03approaches to better understand addiction,
• 01:06psychosis risks, and recovery trajectories.
• 01:10Doctor Ravin's lecture will explore
• 01:12the clinical and neurobiological
• 01:14factors that influence the severity
• 01:16and progression of cannabis withdrawal
• 01:18during periods of abstinence.
• 01:21Using findings from laboratory based
• 01:23cannabis abstinence models,
• 01:25the presentation will examine how
• 01:27individual differences, including biological sex
• 01:30and co use of other
• 01:31substances,
• 01:32may shape withdrawal experiences.
• 01:35The talk will also highlight
• 01:37the role of endocannabinoid
• 01:38system in driving variability in
• 01:40withdrawal trajectories,
• 01:42providing insights into mechanisms that
• 01:44may inform future treatment and
• 01:46intervention strategies
• 01:48for cannabis use disorder.
• 01:50So with that,
• 01:52Rachel, you, have the floor,
• 01:54and I'll,
• 01:55introduce Renato after.
• 01:59Okay. Great. Thank you so
• 02:01much, Romina, for that nice
• 02:02introduction.
• 02:03Can everyone see my slides
• 02:04okay? Maybe just, great.
• 02:07Okay. So, yeah, so I'm
• 02:08really excited to share some
• 02:09of my lab's work on
• 02:11cannabis use
• 02:12and cannabis withdrawal and talk
• 02:14about some of the factors
• 02:16that we've identified that may
• 02:17moderate withdrawal severity as well
• 02:19as duration.
• 02:20And I'll also discuss,
• 02:22several of the underlying mechanisms
• 02:24that we think may contribute
• 02:25to these effects.
• 02:30So I have nothing to
• 02:31disclose.
• 02:32So I think everybody here
• 02:34is well aware that rates
• 02:35of cannabis use are high
• 02:37and increasing
• 02:38alongside the relaxation of cannabis
• 02:40laws.
• 02:42About seven percent of both
• 02:43Canadians and Americans report daily
• 02:46or near daily use of
• 02:47cannabis.
• 02:48And this is really concerning
• 02:49because we know that this
• 02:51pattern of cannabis use is
• 02:52associated with a host of
• 02:54consequences,
• 02:55including the development of a
• 02:57cannabis use disorder.
• 03:00Yet, unfortunately,
• 03:01there remains to be no
• 03:02approved pharmacotherapies
• 03:04to treat cannabis use disorder,
• 03:06and this is despite many
• 03:08candidate medications
• 03:09having been tested.
• 03:12So clearly, there's an urgent
• 03:14need to develop novel and
• 03:15effective treatments for cannabis use
• 03:17disorder, and today I'll talk
• 03:19about couple of ingredients that
• 03:20I think are essential for
• 03:22this process.
• 03:24So the first,
• 03:25we need to identify variables
• 03:26that predict cannabis relapse.
• 03:29We also need to determine
• 03:30factors that moderate this variable.
• 03:33And then we need to
• 03:34map the mechanisms that are
• 03:35driving these moderating effects.
• 03:38So it's already pretty well
• 03:39established that cannabis withdrawal is
• 03:41a robust predictor of cannabis
• 03:43relapse.
• 03:44And just to go over
• 03:45what cannabis withdrawal is, it
• 03:47refers to a constellation of
• 03:49symptoms that emerge following the
• 03:51cessation
• 03:52or the reduction of heavy
• 03:54cannabis use. And it's pretty
• 03:56common up to about ninety
• 03:57percent of people with regular
• 03:59cannabis use will experience withdrawal
• 04:01symptoms.
• 04:03So, these symptoms are clinically
• 04:05significant because we know they
• 04:06interfere with daily functioning.
• 04:09The most common symptoms that
• 04:10we see in our lab
• 04:12tend to be effective symptoms
• 04:13like depression and anxiety or
• 04:15irritability,
• 04:17sleep disturbances
• 04:18as well as appetite disturbances,
• 04:21and also physical symptoms are
• 04:23pretty prevalent.
• 04:26Now these symptoms are pretty
• 04:27distressing,
• 04:28and they often cause people
• 04:30to relapse. And even just
• 04:32the thought of these symptoms
• 04:34can make people
• 04:35relapse during a quit attempt.
• 04:38So, it's important to note
• 04:40that, cannabis withdrawal follows a
• 04:42very distinct and elongated
• 04:44trajectory.
• 04:45So, symptoms typically begin about
• 04:47twenty four hours after someone
• 04:49quits. They usually peak within
• 04:51one week of quitting.
• 04:53And then after about twenty
• 04:54eight days, they pretty much
• 04:55dissipate.
• 04:56And this was first established
• 04:58back in two thousand and
• 04:59three by Budney and colleagues.
• 05:01And about fifteen years later,
• 05:04I was able to replicate,
• 05:06with my lab that as
• 05:07a very similar trajectory.
• 05:12So alongside increasing cannabis use,
• 05:14we're also seeing a rise
• 05:16in tobacco co use in
• 05:18people who use cannabis.
• 05:19And you can see here
• 05:20from this recent study that
• 05:22was published,
• 05:23the blue line here denotes,
• 05:26cannabis and tobacco co use,
• 05:27and you can see that
• 05:28rates are rising.
• 05:30The green line,
• 05:32depicts cannabis exclusive use, so
• 05:34people who just use cannabis
• 05:35alone.
• 05:36And this orange line is
• 05:38the rates of tobacco use,
• 05:39which you can see over
• 05:41the last several decades
• 05:43have immensely declined. But what
• 05:45we're seeing is that in
• 05:47people who use cannabis, we're
• 05:49not seeing that same decline
• 05:50of tobacco use. In fact,
• 05:53people
• 05:54who use cannabis, we're seeing
• 05:55that their rates of tobacco
• 05:57use are actually remaining quite
• 05:58stagnant.
• 06:00And most people who used
• 06:01cannabis actually also
• 06:03co use a tobacco product.
• 06:08So accumulating evidence suggests that
• 06:10tobacco co use is associated
• 06:11with poor treatment outcomes
• 06:14compared to those people who
• 06:15just use cannabis alone.
• 06:17And this has been demonstrated
• 06:18across various studies. So Gray
• 06:20et al showed that people
• 06:21with tobacco co use were
• 06:23fifty percent more likely to
• 06:25relapse during a cannabis treatment
• 06:27trial
• 06:28compared to people with cannabis
• 06:29only use.
• 06:30Moore and Bugney showed that
• 06:32people with current tobacco co
• 06:34use relapsed to cannabis faster
• 06:36than those who previously used
• 06:37tobacco.
• 06:39And Meg Haney's lab at
• 06:40Columbia
• 06:41showed that people with co
• 06:42use were twenty times more
• 06:44likely to relapse to cannabis
• 06:46than people with cannabis only
• 06:48use.
• 06:50So this raises the question,
• 06:51what's responsible for this higher
• 06:53or faster cannabis relapse rates
• 06:56in people with tobacco couse?
• 06:58Could it be withdrawal?
• 07:00So we sought to investigate
• 07:02this relationship looking to see
• 07:03if tobacco co use influences
• 07:06cannabis withdrawal severity
• 07:07in people with a current
• 07:09cannabis use disorder
• 07:10during,
• 07:12an abstinence paradigm where we
• 07:13ask people to quit for
• 07:14twenty eight days.
• 07:16And we chose the twenty
• 07:17eight day period so to
• 07:19coincide with withdrawal trajectory, which
• 07:21we know to be twenty
• 07:22eight days and also with
• 07:25the urinary elimination of THC
• 07:26from the body.
• 07:29So we conducted a secondary
• 07:30analysis. We had,
• 07:33twenty participants
• 07:34who met criteria for a
• 07:35cannabis use disorder,
• 07:36and we parsed them according
• 07:38to their current tobacco use.
• 07:41So we had a group,
• 07:42which we called our co
• 07:43users, and we denoted them
• 07:44by c t plus. So
• 07:46these are people who met
• 07:47for a cannabis use disorder
• 07:49and smoked either ten or
• 07:51more cigarettes per day.
• 07:53And then we had a
• 07:54cannabis only user group,
• 07:56which also,
• 07:58had to include people with
• 07:59very light tobacco use
• 08:01because we actually didn't have
• 08:03enough people who just used
• 08:04cannabis alone. So the CT
• 08:06minus group is a group
• 08:07of individuals who used either
• 08:09cannabis alone
• 08:11or co used but less
• 08:12than five cigarettes per day.
• 08:16And so we just recruited
• 08:17people from the community.
• 08:19Everybody in the sample was
• 08:20male just because it was
• 08:22a secondary analysis.
• 08:24Like I said, they all
• 08:25met for a cannabis use
• 08:26disorder and came up positive
• 08:27for THC at the screening
• 08:29day.
• 08:30They were willing to abstain
• 08:31for cannabis for twenty eight
• 08:33days, but they were not
• 08:34treatment seeking.
• 08:35We excluded anyone with a
• 08:37current psychiatric disorder,
• 08:39if they came up positive
• 08:41for a psychoactive substance other
• 08:42than THC,
• 08:44if they were using any
• 08:45medications that were psychotropic in
• 08:47nature in the last six
• 08:48months,
• 08:49And if they had any
• 08:50serious medical or neurological
• 08:52conditions, we also excluded them.
• 08:55So this was our study
• 08:56design.
• 08:58Once eligibility was confirmed, we
• 09:00asked participants to quit the
• 09:02night before, about twelve hours
• 09:03before coming in for the
• 09:05baseline visit. And this was
• 09:06to try to capture them
• 09:08between the window of acute
• 09:10intoxication, but also before the
• 09:12onset of withdrawal.
• 09:14So then we then got
• 09:15people to come in for
• 09:16weekly study visits where we
• 09:18assess withdrawal. So we use
• 09:20the marijuana withdrawal checklist, which
• 09:22is a self report questionnaire
• 09:24where people
• 09:25rated the severity of fifteen
• 09:27of the most commonly reported
• 09:29withdrawal symptoms, and they were
• 09:31rated on a scale from
• 09:32zero being not present at
• 09:34all to three very severe.
• 09:39And so
• 09:40we also use the timeline
• 09:42follow back just to get
• 09:42a self report measure of
• 09:44substance use.
• 09:45And to get people to
• 09:47achieve absence, we offer them
• 09:49weekly behavioral therapy
• 09:50very briefly.
• 09:52But what really got people
• 09:53to quit is we offered
• 09:54them contingency management.
• 10:03Bonus.
• 10:04And we didn't just rely
• 10:05on self report. We actually
• 10:07collected urine twice weekly. And
• 10:09so we were able to
• 10:10confirm abstinence by chemically.
• 10:14So twenty participants completed the
• 10:16study.
• 10:17We had eleven people in
• 10:18the co use group and
• 10:20nine in
• 10:21the cannabis only co use
• 10:23light group.
• 10:25Two people in the co
• 10:26use group relapsed,
• 10:28and none in the other
• 10:29group relapsed.
• 10:30What was really nice is
• 10:32that the groups were well
• 10:33matched on age,
• 10:35across all parameters of cannabis
• 10:37use, including lifetime years as
• 10:39well as frequency of use.
• 10:41And obviously, they differed on
• 10:43tobacco use. So
• 10:45our co users were smoking
• 10:46about a pack a day
• 10:48and, met for moderate nicotine
• 10:50dependence, while our
• 10:52CT minus group were just
• 10:54using on average about three
• 10:55and a half cigarettes per
• 10:56day
• 10:57and that met for minimal
• 11:00or very mild,
• 11:03nicotine dependence.
• 11:05So when we looked at
• 11:07participants
• 11:08with, cannabis use disorder with
• 11:10either no or low tobacco
• 11:11use,
• 11:12what we saw is this
• 11:13very typical or the classic
• 11:15withdrawal trajectory that we expected
• 11:18to see.
• 11:19However, when we looked at
• 11:20the co use group, we
• 11:22saw a very different pattern.
• 11:23So we saw that in
• 11:24this group, their their severity
• 11:26of cannabis withdrawal was elevated,
• 11:29and it remained elevated across
• 11:31all of the weeks that
• 11:32we assessed withdrawal.
• 11:35And so now that we
• 11:36had a behavior to explain
• 11:37high relapse rates in co
• 11:38users, we were now interested
• 11:40in determining what the molecular
• 11:42mechanisms were that were underlying
• 11:44these elevated rates.
• 11:47And so what we know
• 11:48is that chronic cannabis use
• 11:50alters the endocannabinoid
• 11:51system.
• 11:53Lots of studies showing that
• 11:54chronic use of cannabis causes
• 11:56downregulation
• 11:57of cannabinoid one receptors in
• 11:59the brain. We also see
• 12:00this in the periphery. We
• 12:02see a decrease in endocannabinoid
• 12:04levels that have been measured
• 12:05in blood.
• 12:06And very interestingly, cannabis withdrawal
• 12:09has been linked to endocannabinoid
• 12:11dysregulation,
• 12:12which has been shown by
• 12:13doctor D'Souza,
• 12:15And higher withdrawal severity is
• 12:17actually associated with greater cannabinoid
• 12:19one receptor downregulation.
• 12:22Similar relationships have been shown
• 12:24in the periphery where withdrawal
• 12:26symptoms like depression, anxiety
• 12:28have been linked to endocannabinoid
• 12:30levels.
• 12:31And also studies in both
• 12:32humans and animals have shown
• 12:34that if we use a
• 12:35THA inhibitor,
• 12:38we can increase anandamide levels.
• 12:40So anandamide is one of
• 12:41the prominent endocannabinoids,
• 12:44and thaw is the enzyme
• 12:46that degrades anandmine. So if
• 12:47we block the breakdown
• 12:49of anandamide, we're gonna see
• 12:51increasing levels, and this actually
• 12:53leads to a reduction
• 12:54of cannabis withdrawal symptoms.
• 12:58So what this suggests in
• 12:59the context of our data
• 13:01is that perhaps tobacco co
• 13:02use may exacerbate endocannabinoid
• 13:06dysfunction in people with cannabis
• 13:08use.
• 13:10And so we were able
• 13:11to test this hypothesis
• 13:13in another secondary analysis
• 13:15using data from Romina's,
• 13:18Romina's cohort. So she collected
• 13:20data in people with cannabis
• 13:22use disorder
• 13:24using a validated FAW tracer
• 13:26called CRB.
• 13:28And so just to remind
• 13:29you that FAW degrades the
• 13:31endocannabinoid
• 13:32anandamide,
• 13:33and we used it as
• 13:34a proxy for endocannabinoid
• 13:35activity or signaling.
• 13:37And so the question that
• 13:38we sought to answer was,
• 13:40do fall levels in the
• 13:41brain
• 13:42differ between individuals with co
• 13:44use compared to those with
• 13:46cannabis only use following overnight
• 13:48abstinence?
• 13:51And we also wanted to
• 13:52look to see if tobacco
• 13:54use independently
• 13:55was associated with fall levels.
• 13:58So the sample was, we
• 14:00had five people in the
• 14:01co use group and eight
• 14:03people in the cannabis only
• 14:04group. And just to know
• 14:05that this is a different
• 14:06sample than the one I
• 14:07previously presented before.
• 14:09So these cannabis only users
• 14:11were were really only just
• 14:13using cannabis.
• 14:14And again, we had the
• 14:15groups very well matched on
• 14:17age as well as all
• 14:18the cannabis parameters.
• 14:20And again, they differed on
• 14:21tobacco use, so our CT
• 14:23plus group was using about
• 14:25six cigarettes per day and
• 14:27met for mild,
• 14:28nicotine dependence.
• 14:32So when we looked across
• 14:33regions of interest, and we
• 14:35looked at regions of interest
• 14:37that were that had high
• 14:38levels of cannabinoid one receptors
• 14:40and also had high levels
• 14:42of nicotine acetylcholine
• 14:43receptors.
• 14:45And you can see that
• 14:46across all regions
• 14:48that were examined,
• 14:49the co use group, which
• 14:51is the one denoted in
• 14:52orange, had higher fall levels
• 14:55than people with cannabis only
• 14:57use.
• 14:58And
• 14:59this even though the groups
• 15:01were small,
• 15:02significant differences emerged, in the
• 15:04cerebellum shown here, as well
• 15:06as the substantia nigra.
• 15:09But really this trend was
• 15:10present across all regions examined.
• 15:14And also, when we look
• 15:16to see,
• 15:17how tobacco use or cigarettes
• 15:19per day correlated with FAW,
• 15:21we saw a positive relationship
• 15:23where greater tobacco use actually
• 15:25correlated
• 15:26with higher FAW levels in
• 15:27the cerebellum.
• 15:29And it came out as
• 15:30trend level when we looked
• 15:31at the relationship in the
• 15:33substantia nigra.
• 15:36So what this suggests is
• 15:37that tobacco use may be
• 15:39independently contributing to endocannabinoid
• 15:42dysfunction
• 15:43in people with cannabis use.
• 15:45So just to quickly summarize
• 15:47what I've shown so far
• 15:48is that people who co
• 15:49use have elevated and prolonged
• 15:51withdrawal severity compared to people
• 15:53with cannabis only use. People
• 15:55with co use have higher
• 15:57fall levels than people with
• 15:58cannabis only use.
• 16:00Tobacco use may independently predict
• 16:02greater fall levels,
• 16:04and greater withdrawal severity in
• 16:06co users may actually reflect
• 16:08an exacerbated endocannabinoid
• 16:10dysfunction or dysregulation.
• 16:13And so perhaps the
• 16:14endocannabinoid
• 16:15system is a promising therapeutic
• 16:17target
• 16:18to treat cannabis use disorder
• 16:20as well as people who
• 16:22co use.
• 16:24But one thing I do
• 16:25want to point out is
• 16:26that the endocannabinoid
• 16:27system is not a one
• 16:29size fits all. In fact,
• 16:31there's great variability, and one
• 16:33of theone of the sources
• 16:35of variability comes from sex.
• 16:37So, we see lots of
• 16:38sex differences across the endocannabinoid
• 16:41system.
• 16:42However, our understanding of cannabis
• 16:44use and the endocannabinoid
• 16:46system in females is actually
• 16:48quite limited.
• 16:50So, for people who are
• 16:52doing cannabis use research, I'm
• 16:54sure you're familiar with that
• 16:55females are very underrepresented
• 16:58in cannabis studies.
• 17:00And this is really a
• 17:02problem, especially these days because
• 17:04we're seeing that since legalization,
• 17:07rates of cannabis use in
• 17:08females are actually escalating at
• 17:10a much faster rate in
• 17:11females relative to males.
• 17:14And new data just came
• 17:15out from the states that
• 17:17actually showed that female adolescents
• 17:19now have higher rates of
• 17:21cannabis use
• 17:22compared to male adolescents. So
• 17:25this historical trend of males
• 17:27having higher rates of than
• 17:29females is we're now seeing
• 17:30a reversal
• 17:31in these patterns.
• 17:35And this is really concerning
• 17:36because females actually have worse
• 17:38cannabis treatment outcomes compared to
• 17:40males.
• 17:41So they tend to be
• 17:42less adherent to medications.
• 17:43They have reduced responsiveness
• 17:45to interventions.
• 17:47They often demonstrate higher relapse
• 17:49rates and lower
• 17:51cessation success.
• 17:53And so, again, it could
• 17:54be that cannabis withdrawal may
• 17:54underlie some of these sex
• 17:55specific effects.
• 18:06The data is quite mixed.
• 18:07There's some the data is
• 18:08quite mixed. There are some
• 18:09studies that show that females
• 18:11experience greater withdrawal burden than
• 18:13males,
• 18:14others that show the reverse
• 18:16relationship,
• 18:17and then many studies that
• 18:18actually show that there's no
• 18:20sex differences in cannabis withdrawal
• 18:22severity or
• 18:37And this is quite And
• 18:38this is quite problematic because
• 18:40what we see is that
• 18:41females
• 18:42with cannabis use disorder tend
• 18:44to have higher rates of
• 18:45psychiatric comorbidities,
• 18:48and the symptoms of these
• 18:49comorbidities
• 18:50actually overlap
• 18:51with a lot of the
• 18:52withdrawal symptoms.
• 18:55Second is that the evidence
• 18:57from these studies is derived
• 18:59from
• 18:59mostly studies that just have
• 19:01looked at one time point.
• 19:03They're predominantly
• 19:04cross sectional in nature, and
• 19:06they don't allow us to
• 19:07look at the dynamic
• 19:08or the time varying
• 19:10sex differences that may appear
• 19:12along the trajectory of withdrawal.
• 19:15And I also just wanted
• 19:16to point out that a
• 19:17lot of what we know
• 19:18about the cannabis withdrawal trajectory
• 19:20actually comes from studies
• 19:22that had been predominantly conducted
• 19:24in males.
• 19:26So Gabriela Malamud was a
• 19:28research,
• 19:29was a graduate student in
• 19:30my lab who recently finished
• 19:32her master's
• 19:33and she conducted, some preliminary
• 19:35analysis looking at how sex
• 19:38differences,
• 19:39portray
• 19:40across the cannabis withdrawal trajectory.
• 19:43So, in her project, she
• 19:44investigated how sex influences the
• 19:46severity of cannabis withdrawal symptoms
• 19:49in people with a current
• 19:50cannabis use disorder during twenty
• 19:52eight days of abstinence.
• 19:54And the inclusion exclusion criteria
• 19:56was pretty similar to the
• 19:57one that I showed you
• 19:58from the COUSE study,
• 20:00except here people had to
• 20:01identify as male or female.
• 20:03We also excluded people who
• 20:05smoked more than, ten cigarettes
• 20:07per day as well as
• 20:08females who were pregnant or
• 20:10breastfeeding.
• 20:12So it was the exact
• 20:13same study paradigm as I
• 20:15showed before.
• 20:17And in this study, we
• 20:18had twenty seven participants that
• 20:20completed abstinence,
• 20:21fifteen males, twelve females.
• 20:24Two of the males relapsed,
• 20:25three females relapsed, but there
• 20:27was no sex difference in,
• 20:29the abstinence rate.
• 20:32So the males and females
• 20:34were well matched across all
• 20:36parameters. So, age as well
• 20:38as cannabis
• 20:39parameters and cigarettes and alcohol
• 20:42use per day.
• 20:44But because we had a
• 20:45wide range of recency of
• 20:47use, we controlled for this
• 20:49in our analyses.
• 20:51And so when we looked
• 20:52at males, we saw again
• 20:54that they exhibited
• 20:55the very classic
• 20:57cannabis withdrawal trajectory that has
• 20:58been previously established.
• 21:01And when we looked at
• 21:02females, again, we saw a
• 21:03very different trajectory
• 21:05than males. So females experienced
• 21:08a more prolonged course of
• 21:09withdrawal.
• 21:10And at day twenty eight,
• 21:11they actually had greater withdrawal
• 21:13than males.
• 21:16And so when we first
• 21:17saw this data,
• 21:19I thought perhaps or we
• 21:21theorized that perhaps some of
• 21:22these effects were due to
• 21:23pharmacokinetic
• 21:24sex differences.
• 21:26And we had,
• 21:28data from our urine collection
• 21:30to confirm absence. So we
• 21:31could actually look at carboxy
• 21:33and hydroxy levels, these major
• 21:35metabolites
• 21:36of THC to see if
• 21:37they differed between sex,
• 21:39but they didn't.
• 21:41So perhaps pharmacodynamic
• 21:43effects may be better,
• 21:46suited to explain these sex
• 21:47differences.
• 21:49And,
• 21:50we know that there are
• 21:51sex differences in the endocannabinoid
• 21:53system, although the direction of
• 21:55these effects are really not
• 21:56well established. So, you can
• 21:58find them in both directions
• 22:00in the literature.
• 22:02And we know that chronic
• 22:02cannabis use exerts pharmacodynamic
• 22:05effects, as I explained before.
• 22:06And this has been shown
• 22:07in both males and females.
• 22:09This idea that cannabinoid one
• 22:11receptor downregulation
• 22:12occurs with chronic use.
• 22:14It's also been shown in
• 22:16males and females that,
• 22:18cannabinoid one receptor ability predicts
• 22:20withdrawal severity.
• 22:22But what hasn't been shown
• 22:23in females
• 22:25is that,
• 22:26normalization
• 22:27or upregulation
• 22:29occurs with abstinence. So while
• 22:30this has been
• 22:32documented in two studies in
• 22:33males, it has not yet
• 22:35been examined in females and
• 22:37perhaps,
• 22:39dysregulation in the endocannabinoid
• 22:40system may persist longer in
• 22:43females than males.
• 22:46So another important piece of
• 22:47the puzzle to consider is
• 22:49the role of ovarian hormones.
• 22:51And we know that they
• 22:53influence endocannabinoid
• 22:54system as well as that
• 22:56is it as well as
• 22:57its activity.
• 22:58So they can actually modulate
• 23:00cannabinoid one receptor density or
• 23:02the affinity of these receptors.
• 23:04And these ovarian hormones obviously
• 23:07fluctuate across the menstrual system.
• 23:09So we can actually see
• 23:11across the menstrual system that
• 23:12endocannabinoid
• 23:13activity can
• 23:16change.
• 23:17But one one thing to
• 23:18note is that if hormones
• 23:20do change,
• 23:21across,
• 23:23the with the menstrual cycle,
• 23:25then we would have expected
• 23:26to see some variation
• 23:28in our withdrawal trajectory in
• 23:29females,
• 23:31even though the average,
• 23:33trajectory or severity was stagnant
• 23:36across females.
• 23:37So
• 23:39what we did next was
• 23:40we then looked at the
• 23:41individual trajectories of the females
• 23:43in the study. And when
• 23:45we unpack this data, we
• 23:46actually see that across the
• 23:48nine participants,
• 23:51the data was very heterogeneous.
• 23:53So all of the females
• 23:55looks like that they experienced
• 23:56a very different type of
• 23:58withdrawal trajectory,
• 24:00which suggests that hormones may
• 24:02indeed play a role
• 24:03in the severity of the
• 24:04symptoms.
• 24:05And one thing that we
• 24:06do see that we don't
• 24:07see in males is sort
• 24:08of this later phase elevation
• 24:10or surge in withdrawal severity,
• 24:13that's not present in males.
• 24:15And actually, when you look
• 24:16at males, and these are
• 24:17the thirteen males plotted out,
• 24:19you can actually see that
• 24:21the majority or seventy five
• 24:22percent
• 24:23of males actually showed
• 24:26the expected or the classic
• 24:27withdrawal trajectory. So it wasn't
• 24:29just an average effect, but
• 24:31most of these men are
• 24:32actually showing,
• 24:35this dissipation by day twenty
• 24:37eight and that peak at
• 24:38day seven.
• 24:40And so,
• 24:42what we see just to
• 24:43sum up is that females
• 24:45exhibit
• 24:45quite a different withdrawal trajectory
• 24:47than males. It's a lot
• 24:49more heterogeneous.
• 24:50It seems to be prolonged.
• 24:53These effects are likely mediated
• 24:56by some hormonal changes and
• 24:57endocannabinoid
• 24:58activity across the menstrual cycle.
• 25:02And really just sort of
• 25:04speaks to how much more
• 25:05investigations
• 25:06are needed in this area.
• 25:08So we need to start
• 25:10to look to see if
• 25:11hormones can influence or do
• 25:13influence cannabis withdrawal severity or
• 25:15endocannabinoid
• 25:16activity during abstinence.
• 25:19And, the real clinical relevance
• 25:21comes to speak to if
• 25:23treatments then need to be
• 25:24tailored towards sex.
• 25:27And so that's an area
• 25:29that we're continuing to pursue.
• 25:32So just to sum up,
• 25:34I just wanted to come
• 25:36back to this slide and
• 25:38show that,
• 25:41that we've identified withdrawal as
• 25:43a variable that predicts relapse,
• 25:45that tobacco co use as
• 25:47well as sex, do indeed
• 25:48moderate cannabis withdrawal severity as
• 25:51well as duration.
• 25:52And that the endocannabinoid
• 25:54system may underlie the association
• 25:56between withdrawal and tobacco co
• 25:58use as well as sex.
• 26:01And hopefully, a better understanding
• 26:03of these relationships,
• 26:05can help identify novel therapeutics
• 26:07to overcome some of the
• 26:08barriers that we've had
• 26:10to cannabis treatment innovation.
• 26:14And with that, I'll just
• 26:15thank all of my collaborators,
• 26:18as well as all of
• 26:20my team members,
• 26:22and the funding agencies. And
• 26:24a real special thanks to
• 26:25all the participants
• 26:27who
• 26:28who go through a really
• 26:29intensive,
• 26:30laboratory paradigm to provide this
• 26:32data for us. So
• 26:34thank you all for your
• 26:35attention.
• 26:36And if there's time, I'll
• 26:37be happy to answer any
• 26:39questions.
• 26:41Thank
• 26:41you, Rachel. Great, talk. Maybe
• 26:44we'll have time only for
• 26:45two questions
• 26:47if you answer fast.
• 26:50I think there's a question
• 26:51in the chat.
• 26:53Yeah. Okay.
• 26:54So,
• 26:57do you think, postmenopausal
• 26:59females would have a more
• 27:00uniform,
• 27:02pattern of withdrawal as compared
• 27:04to,
• 27:05well, women not during menopause?
• 27:08Yeah. I mean, I would
• 27:10speculate yes, but until we
• 27:12look at that,
• 27:13we really don't know. So
• 27:15there's just a lot of
• 27:16animal data showing that ovarian
• 27:18hormones
• 27:19influence endocannabinoid
• 27:21activity,
• 27:22but there's really not a
• 27:23lot of good studies in
• 27:24humans showing what happens
• 27:26premenopausal
• 27:27or
• 27:28after menopause
• 27:29and even,
• 27:31while while while females are
• 27:33menstruating.
• 27:35Thank you. I have one
• 27:36question. And the last question,
• 27:38do you think there are
• 27:39specific times within the menstrual
• 27:41cycle that would predict better,
• 27:46success
• 27:47rate? Yeah. So we couldn't
• 27:48look at that from our
• 27:49data. In fact,
• 27:52we didn't have a lot
• 27:52of women on contraceptives, and,
• 27:56females are not very good
• 27:57at recalling,
• 27:59when the last day of
• 28:00the the first day of
• 28:01their last menstrual cycle was.
• 28:03So it was really hard
• 28:03to track
• 28:04where in their cycle they
• 28:05were during the
• 28:07study, but there is some
• 28:08evidence from the tobacco literature
• 28:10suggesting that,
• 28:12being in the early phase
• 28:13or late phase when estrogen
• 28:15levels are high or low,
• 28:16I mean, there's mixed evidence,
• 28:18may be beneficial and produce
• 28:20greater treatment success.
• 28:22Which phase that is, we
• 28:24don't know yet. And whether
• 28:25that's the same as the
• 28:26tobacco field or it's specific
• 28:27for cannabis, we also don't
• 28:29know.
• 28:30So, like I said, we're
• 28:31just sort of scratching the
• 28:32surface here, and I think
• 28:33there there remains a lot
• 28:35of unanswered questions.
• 28:40Okay.
• 28:41I guess just one comment.
• 28:43One question. Do you think
• 28:44the tarpine,
• 28:45tarpine concentration,
• 28:47may have explained some of
• 28:48your findings?
• 28:51Yeah. I mean, cannabis is
• 28:53just such a complex plan,
• 28:55so we haven't really gotten
• 28:57into exactly what people are
• 28:59using. We're really trying to
• 29:00do a good job to
• 29:01try to document
• 29:03which strains,
• 29:04the potency.
• 29:06We're like, in Quebec, everything
• 29:08is really well regulated. So
• 29:09we're trying to
• 29:10get pictures of the products
• 29:12people are using, which sometimes
• 29:14lists the terpenes,
• 29:15but not always the concentrations
• 29:17of them. So
• 29:19so we're trying, but this
• 29:20is a really it's really
• 29:22hard to figure out exactly
• 29:23what people are using.
• 29:26Yeah. And I guess just
• 29:26the last comment from, doctor
• 29:28De Souza that,
• 29:30of course, c v one
• 29:31receptor availability
• 29:33is lower in
• 29:34tobacco smokers, which I'm sure
• 29:36you know. And, however,
• 29:38I don't think they're lower
• 29:39in alcohol
• 29:40use disorder,
• 29:42or at least not consistently.
• 29:44And I'm wondering with the
• 29:46last question whether you've looked
• 29:47into alcohol use.
• 29:49Yeah. I mean, our we
• 29:50exclude anyone with an alcohol
• 29:52use disorder.
• 29:53And, I mean,
• 29:55people
• 29:56are drinking recreationally, but the
• 29:58levels are very, very low.
• 30:00So,
• 30:01I think that the average
• 30:02works out to less than
• 30:03a drink a day.
• 30:05So we could look at
• 30:06that, but I again, because
• 30:08the the volume that people
• 30:09are consuming are quite low,
• 30:10I wouldn't expect in this
• 30:11sample for it it to
• 30:12have an effect.
• 30:14Excellent.
• 30:15Well, thank you very, very
• 30:16much, Rachel. Great talk.
• 30:18And,
• 30:19with that, we are going
• 30:20to be moving,
• 30:22on to Renato.
• 30:24Doctor Polimanti's
• 30:25research focuses on applying big
• 30:27data analytics, genomics, and computational
• 30:30biology to better understand the
• 30:32biological
• 30:33and epidemiological
• 30:34mechanisms
• 30:35underlying neuropsychiatric
• 30:37disorders,
• 30:38substance use, and complex behavioral
• 30:40traits.
• 30:41Doctor Polimanti
• 30:42is one of the recipients
• 30:43of the Yale Cannabis Research
• 30:45Center's pilot project award.
• 30:47His lecture will explore the
• 30:49neurobiological
• 30:50mechanism
• 30:51that distinguish
• 30:52the cannabis use from cannabis
• 30:54use from cannabis use disorder
• 30:56through the integration of genomics
• 30:58and brain imaging data. The
• 31:00presentation will highlight how listing
• 31:03patterns of brain connectivity
• 31:05and genetic architecture
• 31:07are associated with cannabis and
• 31:09cannabis use disorder,
• 31:10particularly within networks involved in
• 31:12cognitive control,
• 31:14salience processing, and default mode,
• 31:17functioning.
• 31:18Findings will also examine the
• 31:19role of neurodevelopmental,
• 31:21inflammatory, and immune related pathways
• 31:24in the progression from cannabis
• 31:25use to problematic use.
• 31:28Additionally, the talk will discuss
• 31:30how these discoveries may inform
• 31:32future therapeutic
• 31:33strategies, including the identification
• 31:36of potential drug repurposing candidates
• 31:39targeting pathways linked to COD.
• 31:42Doctor Bolimanti,
• 31:44the floor is yours.
• 31:46Great.
• 31:48Thank you so much for
• 31:49the invite. I'm very happy
• 31:50to present this project. And
• 31:52now I should be in
• 31:53presentation mode,
• 31:55And so I'm not seeing
• 31:56anyone right now. Great. Okay.
• 31:59So,
• 32:00this is the study that
• 32:01I'm going to present.
• 32:03And, again, here are my
• 32:05competing interest, which are not
• 32:07related to cannabis
• 32:08research.
• 32:09And, this is my group.
• 32:11Usually, I like to start
• 32:12with with their picture,
• 32:14also because, again, they do
• 32:15most of the work. In
• 32:16this case, like, most of
• 32:17the analysis that I'm going
• 32:19to present were done by
• 32:20Rapunzel Chen.
• 32:22And so, like, as mentioned,
• 32:24like, this study was,
• 32:26funded by a pilot award,
• 32:28which, I'm very grateful for.
• 32:31And,
• 32:32so before getting started, since
• 32:34I know how diverse
• 32:35is the audience attending these
• 32:37webinars,
• 32:38I would like to give
• 32:39you, like, some,
• 32:41basic concept about
• 32:43human genetic research, which is
• 32:45the,
• 32:46base of the study that,
• 32:49we did.
• 32:50And so, like, our starting
• 32:52step are genome wide association
• 32:54studies.
• 32:55These are brute force experiments
• 32:57and why they are, why
• 32:59I'm calling them like this
• 33:00is because we are testing
• 33:02all the variants across the
• 33:04genome to find the
• 33:06associations. And what this means,
• 33:07it means that we are
• 33:08finding for variants that are
• 33:10more frequent.
• 33:11For example, if exact case
• 33:13control definition, more frequent in
• 33:14people with the disease than
• 33:15people without. In this case,
• 33:17it would be the SNP
• 33:19number three where you see
• 33:20that there are more,
• 33:23allele
• 33:23in the case group than
• 33:25in the control group.
• 33:27And, again, like, this is
• 33:28a technology that now has,
• 33:30I think, almost twenty years
• 33:33of age.
• 33:34And, in twenty years,
• 33:36more than seventy thousand
• 33:38study have been published, and
• 33:40more than one million
• 33:41association
• 33:42have been reported.
• 33:44Here is a a plot
• 33:46from the GWAS catalog for
• 33:47people that work in human
• 33:48genetics. This is,
• 33:50something that people showed all
• 33:51the time, but, again, it's
• 33:53it's all looks so what
• 33:54what you see here is
• 33:56the chromosome, which is are
• 33:57these tiny
• 33:58streams, and then all these
• 34:00dots are genetic associations. So
• 34:02you can see here that,
• 34:04variants are associated with complex
• 34:06traits are everywhere.
• 34:08And so, like, because the
• 34:10architecture of common traits and
• 34:12disease
• 34:13is due to,
• 34:14the contribution of many variants
• 34:16with very small
• 34:18effect size. And with respect
• 34:20to,
• 34:21what we can do with
• 34:23this kind of information,
• 34:24we can use this genetic
• 34:26association
• 34:27to investigate both the epidemiology
• 34:29and then biology
• 34:30of human traits and disease.
• 34:32This is another
• 34:33very famous plot in, in
• 34:35genetic research. It's a Manhattan
• 34:37plot. And here on the
• 34:38x axis, you have, like,
• 34:40our chromosomes.
• 34:41And then here, you have,
• 34:42like, a statistical significance.
• 34:45The highest
• 34:46the the the dot here,
• 34:47the strongest is the statistical
• 34:49significance, and this each dot
• 34:51is a is a variant.
• 34:53And so depending on how
• 34:54many variants we get,
• 34:56surviving our multiple testing correction,
• 34:58which is this banded line,
• 35:01We can detect, like, information
• 35:03about which are the pathways
• 35:05involved in our disease of
• 35:06interest. We can create a
• 35:08polygenic risk score to distinguish
• 35:10case and controls.
• 35:12We can perform a causal
• 35:13inference analysis.
• 35:15We can integrate
• 35:16other omic domains to understand
• 35:19how a genetic variance can
• 35:20lead to the disease
• 35:22through which molecular mechanism
• 35:24in the brain or in
• 35:25other organs,
• 35:27can affect the disease risk
• 35:29or conduct multivariable analysis to
• 35:32understand
• 35:32how
• 35:33two traits, two different diseases
• 35:36that we observe
• 35:37as comorbid
• 35:38share genetic,
• 35:40factors.
• 35:41So,
• 35:42with respect to cannabis research,
• 35:44here are the,
• 35:45five
• 35:46large scale
• 35:48GWAS published in the last
• 35:50few years. And here you
• 35:51can see that we have
• 35:52three for cannabis use and
• 35:54two for, cannabis use disorder.
• 35:57And so, like, again, several
• 35:59of these,
• 36:00saw, like, Joel Scalenta Group
• 36:02as the primary,
• 36:04leader in these studies. And
• 36:06so what we can do
• 36:07with this information?
• 36:08With this information, we can
• 36:10understand the difference
• 36:12between
• 36:13cannabis use and cannabis use
• 36:14disorder,
• 36:15how strong is the genetic
• 36:17overlap between these two, and
• 36:19how these two
• 36:20differ in their relationship
• 36:22with other,
• 36:23psychiatric
• 36:24and behavioral
• 36:25traits. And so this is
• 36:27another
• 36:28plot from one of the
• 36:29study. Again, it's one of
• 36:30the study which was led
• 36:32by, Joel Skeleton group. And
• 36:34you can see here that
• 36:35we have multiple,
• 36:37secondary disorders
• 36:39and other behavioral
• 36:40phenotypes.
• 36:41And so this is, an
• 36:42analysis called a genomic structural
• 36:44equation modeling. And so it
• 36:46tries to converge,
• 36:48con
• 36:49converge, like, genetic correlation across
• 36:52these traits
• 36:53in latent factors that underline
• 36:56the genetic
• 36:58relationship among these these different
• 37:00conditions. And so here I
• 37:01highlighted the cannabis use disorder
• 37:04and then cannabis use. You
• 37:05can see that cannabis use
• 37:07disorder cluster together other substance
• 37:09use disorder,
• 37:10opioid use disorder, alcohol use
• 37:12disorder.
• 37:14And we start away from
• 37:15cannabis use, and cannabis use
• 37:17cluster
• 37:18with other traits, like a
• 37:20number of sexual partner,
• 37:22smoke initiation,
• 37:24Townsend Deprivation Inserts, which is
• 37:26a material deprivation, and they
• 37:28cluster on this impulsivity,
• 37:30a risk taking,
• 37:32factor. And so you can
• 37:34see here that these two
• 37:35traits
• 37:36are partially,
• 37:38genetically,
• 37:39correlated. And so here here
• 37:41you can see, like, the
• 37:42genetic correlation between the impulsivity
• 37:44risk taking factor and the
• 37:45substance dependence
• 37:47factor, which is about six
• 37:49point sixty five. And so
• 37:50you can see how different
• 37:51they are. And then we
• 37:53can look up to our
• 37:54recent preprint
• 37:55where they increased
• 37:58the the study
• 37:59of,
• 38:01different cannabis trait, including
• 38:03adverse use of cannabis,
• 38:05frequency
• 38:06of use of cannabis, and
• 38:08cannabis use disorder. And, yeah,
• 38:09you can see, like, different
• 38:11trends of this relationship. As
• 38:13I mentioned, this is the
• 38:14for example, looking to a
• 38:16condition and
• 38:18social,
• 38:19determinants of health, we have
• 38:20material deprivation here, and we
• 38:22can see that there is
• 38:24a positive genetic correlation, which
• 38:26is stronger for cannabis use
• 38:27disorder and less stronger for
• 38:29cannabis use,
• 38:31ever
• 38:32cannabis use, and frequency is
• 38:34kind of in the middle.
• 38:35Most of the time, we
• 38:36we can see that frequency
• 38:38is
• 38:39somehow in the middle between
• 38:41the fact that we see
• 38:42for cannabis use disorder and
• 38:43the fact that we see
• 38:44for ever using cannabis.
• 38:47But, okay, in some cases,
• 38:48like, we can see, like,
• 38:49something that is reversed. For
• 38:51example, for working memory, we
• 38:53see, like, a positive genetic
• 38:55correlation, which is lower
• 38:57for cannabis use disorder, but
• 38:59is higher for frequency and
• 39:01ever use.
• 39:03Verbal reasoning, we see opposite
• 39:05collection for ever use and
• 39:07cannabis use disorder. And so,
• 39:08like, similar pattern also for
• 39:10IQ, educational attainment,
• 39:12income. Again, childhood IQ, we
• 39:15see, like, it's probably due
• 39:17to less power, but we
• 39:18see a similar trend. Reaction
• 39:20time, the effect is basically
• 39:21null. With respect to mental
• 39:24health outcomes,
• 39:26we can see a similar
• 39:27trend where the cannabis use
• 39:29disorder is most strongly,
• 39:32genetically correlated
• 39:33with a psychiatric
• 39:35condition and also, like, negative
• 39:38mental health outcomes, but we
• 39:40have some exceptions. For example,
• 39:42with the autism and the
• 39:43Noroxera nervosa, it's actually
• 39:45ever cannabis use that has
• 39:47stronger genetic relation. So you
• 39:48we we can see how
• 39:49these two traits,
• 39:51are different, and they're different
• 39:53in their relationship with
• 39:55other
• 39:56brain related outcomes. And here,
• 39:58we can see, like, some
• 39:59substance use,
• 40:01use disorder,
• 40:02phenotypes, and when we look
• 40:04for dependence,
• 40:06and,
• 40:07other
• 40:08strongly related,
• 40:10behavioral,
• 40:11linked to addiction, we can
• 40:12see that cannabis use disorder
• 40:14is the most strongly related.
• 40:17And so, like,
• 40:19our goal in this study
• 40:20was to try to understand
• 40:21better the
• 40:23brain biology
• 40:24underlying this difference between cannabis
• 40:26use disorder and cannabis use.
• 40:28And to do that, we
• 40:29use the UK Biobank. This
• 40:31is a a large quarter
• 40:33that enrolled more than five
• 40:35hundred thousand people in the
• 40:37UK, and they generated the,
• 40:39brain multimodal brain imaging data,
• 40:42for for
• 40:43a large chunk of this
• 40:44population. In twenty twenty five,
• 40:46the UK Biobank released
• 40:49brain,
• 40:51body, and bone scans for
• 40:53one hundred thousand volunteers.
• 40:54In the analysis that I'm
• 40:55presenting today, we used
• 40:58a previous release of the
• 40:59data, which included thirty three
• 41:01thousand participants
• 41:03and almost four thousand brain
• 41:05imaging derived phenotypes
• 41:07derived from six different brain
• 41:09modalities that can give us
• 41:10information about,
• 41:12variation
• 41:13in brain structure and function.
• 41:17And yet we report like
• 41:18this, this plot about the
• 41:21genetic component
• 41:22of these different,
• 41:24brain imaging phenotype. And so
• 41:25we have genetic information,
• 41:28regarding cannabis use disorder, cannabis
• 41:30use, and genetic information about
• 41:32these brain imaging phenotypes.
• 41:35In a previous study,
• 41:36which was based mostly on
• 41:38observational,
• 41:40analysis,
• 41:42the was the UK Biobank
• 41:44was used to understand the
• 41:45difference between,
• 41:47cannabis users and controls. And
• 41:49you can see here that
• 41:50the sample size is much
• 41:52smaller because the this study
• 41:54focused mostly on
• 41:56individuals that have actual information
• 41:58about
• 41:59cannabis use. And in this
• 42:01study, they identified that people
• 42:03with,
• 42:04that use cannabis in their
• 42:05life have lower,
• 42:07white matter integrity,
• 42:09higher immunity
• 42:11in the corpus callosus,
• 42:13and reduced connectivity
• 42:15in the default mode and
• 42:16the same internal executive networks.
• 42:19And so, like, in this
• 42:20study,
• 42:21they use
• 42:22individual level information to try
• 42:24to understand the association between
• 42:26brain variation
• 42:27and cannabis use. They were
• 42:29not able to do an
• 42:30analysis for cannabis use disorder
• 42:32because cannabis use disorder
• 42:34doesn't have a high prevalence
• 42:35in the UK Biobank. We
• 42:36don't have that that information.
• 42:38And so in the study
• 42:41that was funded by this
• 42:42pilot award,
• 42:43instead of using information about
• 42:45individual participants,
• 42:47we modeled
• 42:48genetic effects.
• 42:50Genetic effects,
• 42:51for related to cannabis use
• 42:53disorder, to cannabis use, and
• 42:55to this four thousand brain
• 42:57imaging phenotypes.
• 42:59And our first goal was
• 43:00to understand if there is
• 43:02a different relationship
• 43:03between
• 43:04the genetic effects
• 43:06linking cannabis use disorder to
• 43:08brain variation
• 43:09and those related to cannabis
• 43:10use.
• 43:11And then to follow-up this
• 43:13first analysis, we want to
• 43:15understand if the biology
• 43:17underlying this
• 43:18pleiotropic effect,
• 43:20effect that are shared between
• 43:21a brain variation
• 43:22and this cannabis phenotype
• 43:25could be targeted by existing
• 43:27molecular compounds.
• 43:30And here are some, initial,
• 43:32results. And so, like, here,
• 43:34we can see, like, some
• 43:35statistics
• 43:36about global genetic correlation. So
• 43:39correlation
• 43:40across,
• 43:41genetic factor in the whole
• 43:43genome. And here we can
• 43:44see that
• 43:46several of these are related
• 43:47to functional connectivity
• 43:49and mostly relate to the
• 43:51default
• 43:52mode network.
• 43:54We also see,
• 43:56like, a white matter,
• 43:58microstructure
• 43:59association
• 44:00in the right superior thalamic
• 44:02region. And but the important
• 44:04part here is that this
• 44:05effect
• 44:06appears to be specific
• 44:08with only one of the
• 44:09two cannabis phenotype. So they
• 44:11don't appear to be shared.
• 44:13And so, like, we can
• 44:14see, for example, like, for
• 44:15this first brain function, the
• 44:17effect is specific to cannabis
• 44:19use disorder, but it's not
• 44:20present for cannabis use. And
• 44:22the same things follow the
• 44:23others. And in the first
• 44:25and in all these cases,
• 44:26we can see that the
• 44:27statistical
• 44:28difference between the cannabis use
• 44:30versus cannabis use disorder or
• 44:32genetic correlation aspect is statistically
• 44:34significant.
• 44:35But, again, this is only
• 44:36the the first step of
• 44:37our analysis. We wanted to
• 44:39go deeper. And so what
• 44:41and here is, like, again,
• 44:43like,
• 44:44a a plot about one
• 44:46of these,
• 44:47functional connectivity,
• 44:49phenotypes that we identified.
• 44:51And this is particularly related
• 44:52to the fourth mode network.
• 44:54And, again, if we compare
• 44:57the, genetic correlation between cannabis
• 44:59use and cannabis use disorder,
• 45:00we can see that there
• 45:01is a strong and statistically
• 45:03significant
• 45:04relationship with cannabis use, but
• 45:06not for cannabis use disorder.
• 45:08But as I was mentioning,
• 45:09we wanted to do, a
• 45:11deeper analysis. And so we
• 45:13moved from,
• 45:14a genetic correlation analysis
• 45:16to a latent causal variable
• 45:18analysis.
• 45:18So in this approach, we
• 45:20try to identify
• 45:21if the genetic effect
• 45:24that is shared
• 45:25between two different phenotypes,
• 45:28in this case, between
• 45:30cannabis phenotype and a brain
• 45:32imaging derived,
• 45:33phenotype
• 45:34is due to a latent
• 45:35causal variable. And if this
• 45:37causal variable has a full
• 45:39causal effect,
• 45:40it could negate partial,
• 45:42partially this this effect.
• 45:44And so we run this
• 45:45analysis, and we were able
• 45:46to identify additional,
• 45:50brain imaging phenotypes. And also
• 45:52in this case,
• 45:53what we observed is that
• 45:56each of these,
• 45:58relationship
• 45:58was specific to only one
• 46:00of the cannabis phenotype.
• 46:02So we found,
• 46:03genetic relationship
• 46:05specific
• 46:06to cannabis use and genetic
• 46:08relationship
• 46:09specific to cannabis use disorder.
• 46:11Also, in this case, there
• 46:12was,
• 46:13an overrepresentation
• 46:14of,
• 46:16functional connectivity
• 46:17phenotypes. But we also observed,
• 46:20like,
• 46:21cortical thickness
• 46:22and,
• 46:23again, white matter,
• 46:26microstructure,
• 46:27phenotypes,
• 46:28again, specifically related to cannabis
• 46:30use disorder.
• 46:32Also in this case,
• 46:33again, I want to show
• 46:35one of these brain imaging
• 46:37brain imaging brain imaging phenotype
• 46:39and specifically one, again, related
• 46:40to the default,
• 46:42model network. And in this
• 46:43case, you can see how
• 46:45stronger is the relationship with
• 46:47cannabis use disorder while it's
• 46:49null for the cannabis use
• 46:51disorder,
• 46:52phenotype.
• 46:55And here is, again, summarizing,
• 46:57our different genetic relationship that
• 46:59we observed
• 47:01with respect to the connectivity
• 47:03between different networks.
• 47:04And so, like, we have
• 47:06different,
• 47:07nodes and the edge that
• 47:08we identified.
• 47:10And so, like, again, color
• 47:12coding based on if these
• 47:14were genetic
• 47:15correlation
• 47:16or if these
• 47:17were,
• 47:18genetic
• 47:19genetically causal proportion. So potential
• 47:22causal relationship within,
• 47:25the BRAIN imaging phenotype and
• 47:27one of the phenotype. As
• 47:28you can see, as I
• 47:29mentioned, like, we observe both
• 47:32differences
• 47:32between the two, cannabis phenotype,
• 47:34cannabis use versus cannabis use
• 47:37disorder. But, also, we observe
• 47:39that there are differences between,
• 47:40like, the
• 47:42genetic,
• 47:44correlation,
• 47:45the phenotypes identified by the
• 47:46genetic correlation and the phenotypes
• 47:48identified
• 47:49by the Ladd and Kousser
• 47:50variables. So those phenotype with
• 47:52genetically Kousser proportion.
• 47:55And so to follow-up even
• 47:57more,
• 47:59this analysis,
• 48:00we decided to do a
• 48:01local genetic correlation analysis. In
• 48:04particular, we want to do
• 48:05this because,
• 48:06to follow-up on the possible
• 48:08causal relationship identified in our
• 48:11study, we also performed
• 48:12Mendelian randomization approach. This is
• 48:15another,
• 48:16genetically informed causal inference method,
• 48:19and we didn't find consistent
• 48:20between the latent causal variable
• 48:22analysis
• 48:23and the,
• 48:24Mendelian randomization results.
• 48:27So we politicized
• 48:28that there may be specific
• 48:30loci
• 48:31that
• 48:32are linking
• 48:33brain variation with one of
• 48:35these gamma disc phenotype.
• 48:37And we identified
• 48:38five different region in the
• 48:40genome where we observed
• 48:42this local genetic correlation.
• 48:45In this case, again, if
• 48:46we consider only results that
• 48:48survive,
• 48:49false discovery rate, multiple testing
• 48:51correction, we can see that
• 48:52there is some specificity.
• 48:54But in this case, we
• 48:55also identified that there may
• 48:57be some overlap
• 48:59that link the same region,
• 49:01to the other cannabis phenotype
• 49:04in relation to one other
• 49:06of the functional connectivity phenotype.
• 49:08Also, one thing that I
• 49:09want to highlight that
• 49:11the brain imaging phenotype that
• 49:13we identified,
• 49:15in the local genetic correlation
• 49:17analysis were all related to
• 49:18functional connectivity.
• 49:20And some of these,
• 49:21were
• 49:23interesting with respect to brain
• 49:24biology.
• 49:26For example,
• 49:27in in some in for
• 49:30the ENPP
• 49:31six gene
• 49:32is involved in the,
• 49:34neurogenesis.
• 49:36The
• 49:38gene is involved in axon
• 49:41guidance.
• 49:42The SMHC
• 49:43two,
• 49:44gene is involved in neurogenesis.
• 49:48The LHPP
• 49:49and the FAM fifty three
• 49:51b gene are both genes
• 49:52that were previously identified in
• 49:55large scale,
• 49:56genome wide analysis of other,
• 49:59substance use disorder. For example,
• 50:01the FAM fifty two b
• 50:03gene was identified in a
• 50:04coven dependence
• 50:06GWAS.
• 50:07The LHPP
• 50:09gene was identified
• 50:10in a study investigating risk
• 50:12taking behavior
• 50:13with respect in the context
• 50:14of,
• 50:15alcohol dependence.
• 50:17But again, like,
• 50:19here, we can see there
• 50:20are many other genes. And
• 50:21so, like, there may be
• 50:22other kind of, like,
• 50:25dynamics
• 50:25that go beyond
• 50:27specific brain biology
• 50:29that could link
• 50:30brain variation
• 50:32to, one of these kind
• 50:33of this phenotype.
• 50:35And so,
• 50:37we try to do also
• 50:38a colocalization
• 50:39analysis to identify if there
• 50:41was a specific variant in
• 50:43this region
• 50:44that was adding causal effect
• 50:46on both
• 50:47brain imaging variation
• 50:50and the cannabis phenotype, and
• 50:51we didn't see that. So
• 50:53there wasn't a single causal
• 50:54variant
• 50:55underlying
• 50:56this local genetic correlation.
• 50:59We also didn't see a
• 51:00mediate petrioplaiotomy.
• 51:02At least, we didn't see
• 51:03strong evidence of it because
• 51:04of the, null results that
• 51:07we observe in the Mendelian
• 51:08normalization.
• 51:09In this case, we don't
• 51:11expect,
• 51:12to, be biased by misclassification
• 51:16because, actually, we are observing
• 51:17two different dynamic between cannabis
• 51:19use and cannabis use disorder.
• 51:21So the misclassification
• 51:23should have increased the likelihood
• 51:25to see,
• 51:26convergence between two different phenotype.
• 51:29And so what we expect
• 51:31is that pleiotropic mechanism
• 51:33linking brain variation
• 51:35to cannabis,
• 51:37phenotypes
• 51:38should be, due to different
• 51:40type of a pleiotropy. To
• 51:41a pleiotropy where
• 51:43a variance located
• 51:45in the
• 51:46same genes or in two
• 51:48different genes
• 51:49in strong link and disequilibrium
• 51:52may be,
• 51:53responsible for the pleiotropy that
• 51:55we are observing. And so
• 51:57to further investigate this,
• 51:59we did a gene ontology
• 52:01analysis. And so what are
• 52:03gene ontologies?
• 52:04Gene ontologies are is a
• 52:05classification
• 52:06to categorize,
• 52:08loci depending on,
• 52:11which processes
• 52:12they play a role in,
• 52:14where
• 52:15the gene products are expressed,
• 52:17or what type of action
• 52:18they do. And so, like,
• 52:19there are three main categories,
• 52:21biological processes, molecular function, and
• 52:24cellular components.
• 52:25And so we run these
• 52:27gene ontology analysis,
• 52:29identified,
• 52:30more than five hundred
• 52:32gene ontology terms surviving
• 52:34a strict Bonferroni,
• 52:37significance.
• 52:37And these were shared across
• 52:40cannabis use, cannabis use disorder,
• 52:43and the pleiotropic brain imaging
• 52:45phenotype. And here is the
• 52:47distribution across the three different
• 52:48categories,
• 52:49biological processes, molecular function, and
• 52:52cellular components.
• 52:53Of course, in these five
• 52:54hundred
• 52:56geotems,
• 52:56we identified
• 52:58several
• 52:59that were related to brain
• 53:00biology.
• 53:01But the things that was
• 53:03interesting is that the, top
• 53:06terms,
• 53:07for both,
• 53:08cannabis use and cannabis use
• 53:10disorder
• 53:11were related to immune functions.
• 53:13Specifically, for cannabis use, the
• 53:15top
• 53:17genotype terms were related to
• 53:19regulation of inflammatory
• 53:21response
• 53:21and cell activation
• 53:23involved in immune response.
• 53:25And for cannabis use disorder,
• 53:27the top terms were apoptotic
• 53:29signaling pathway and leukocyte differentiation.
• 53:32We also identified
• 53:34some gene ontology that were
• 53:35specific
• 53:36to cannabis use and the
• 53:38pleiotropic
• 53:39brain imaging derived phenotypes, but
• 53:41but they were not shared
• 53:42with cannabis use disorder. And
• 53:44in this case, we found,
• 53:45like, things that are linked
• 53:47to brain biology and in
• 53:48particularly about cell communication
• 53:50and
• 53:51synapse
• 53:52assembly.
• 53:54And so like what we
• 53:55want to do next is
• 53:57to try to understand if
• 53:58these
• 53:59molecular pathways and molecular function,
• 54:01biological processes, and cellular component
• 54:04could give us some insight
• 54:06about
• 54:07molecular compounds that could target
• 54:11this,
• 54:12pleiotropic mechanism linking brain variation
• 54:14with canary spenotype. So what
• 54:16we did was
• 54:17a genetically inform drug report
• 54:19processing analysis. And specifically, what
• 54:22we did
• 54:23was
• 54:24using these
• 54:25pathways and
• 54:26try to identify if,
• 54:30non molecular compounds and in
• 54:32particular, their transcriptomic profiles was
• 54:35somehow,
• 54:36statistically matching what we were
• 54:38observing with respect to,
• 54:40the genetically
• 54:42regulated mechanism led to this
• 54:44plaiotomy.
• 54:45And this is what we
• 54:46found. And so we identified
• 54:48different compounds,
• 54:50and, we did the analysis
• 54:52stratified by different type of
• 54:53gene ontologies. And so we
• 54:55did, one analysis for biological
• 54:57processes, one analysis for molecular
• 54:59function, and one analysis for
• 55:01cellular components.
• 55:02So, of course,
• 55:04there is some kind of,
• 55:05like, relationship
• 55:06across these three these three
• 55:08different,
• 55:09categories, so they are not
• 55:11completely independent.
• 55:13But the fact that for
• 55:14most of the genes, we
• 55:16observe consistent evidence
• 55:18across different gene ontology category
• 55:21further support the reliability
• 55:23of the results.
• 55:24Interestingly,
• 55:26we identified different type of
• 55:28of of drugs. And so
• 55:30for example, the,
• 55:32methylated benzatholium
• 55:33chloride
• 55:34is an antiseptic.
• 55:36The zinc cocaine
• 55:38is an anesthetics.
• 55:40And so, like, there are,
• 55:42Danazole
• 55:43and Raloxifen
• 55:45that are in.
• 55:46Modulator,
• 55:47there is a
• 55:49antidepressant
• 55:50that is con currently was,
• 55:52recently,
• 55:53discontinued,
• 55:54maprotrelin.
• 55:56And so, like, you can
• 55:57see there are different kind
• 55:58of, like,
• 56:00molecular compounds
• 56:02that appears to target
• 56:04this pleiotropic mechanism that we
• 56:05identified between canopy sphenotype
• 56:08and brain variation.
• 56:10And so, like, here is
• 56:12kind of, like, a summary
• 56:14of the
• 56:15key method key findings from
• 56:17this study. We identified that,
• 56:20there are distinct brain connectivity
• 56:22patterns,
• 56:23related to cannabis use versus
• 56:25cannabis use disorder.
• 56:27The pleiotropy of this cannabis
• 56:29phenotype with brain variation doesn't
• 56:31seem to be due to
• 56:32causal relationship,
• 56:33but actually, it could be
• 56:35due to local genetic correlation
• 56:37driven by shared pathway and
• 56:39not by specific causal variants
• 56:41that have effect on both
• 56:43brain variation and cannabis phenotypes.
• 56:46We saw an arrangement for
• 56:48immune pathways, and so this
• 56:50may open new direction in
• 56:51cannabis
• 56:52research. And, also, we found
• 56:54some evidence
• 56:55supporting
• 56:56that brain imaging
• 56:58integrated in brain imaging with
• 57:00a genetically informed drug repurposing
• 57:03analysis
• 57:04could uncover
• 57:05new dark targets for for
• 57:07cannabis,
• 57:08treatments.
• 57:09And with respect to what
• 57:11we are doing now,
• 57:13currently, we are,
• 57:14working on the much larger
• 57:16expanded
• 57:17UK Biobank brain imaging. So
• 57:19moving from thirty three thousand
• 57:21individuals to one hundred thousand
• 57:23individuals to understand better
• 57:25the relationship of the polygenic
• 57:27risk of cannabis use disorder
• 57:29and cannabis use on different
• 57:32brain wide,
• 57:34patterns. And so, like, the
• 57:35goal is to identify
• 57:37a mechanism
• 57:38that are
• 57:39both
• 57:40specific
• 57:41for one of these cannabis
• 57:42phenotype, but also those that
• 57:44are shared. And because of
• 57:45the larger sample size, we
• 57:46also would be able to
• 57:48investigate
• 57:48better,
• 57:50potential sex differences, which we
• 57:52didn't do in this initial
• 57:54study. Another aspect that,
• 57:56we would like to expand
• 57:58is to understand
• 58:00how single,
• 58:02cell specific mechanism
• 58:04may play a role
• 58:05in linking,
• 58:07brain variation to cannabis use
• 58:09disorder and cannabis use. And
• 58:11so we are going to
• 58:12perform a transcriptome
• 58:14wide analysis
• 58:15to understand
• 58:16how,
• 58:17genetically regulated transcriptomic changes
• 58:19can be associated
• 58:21to brain variation and to
• 58:23cannabis phenotype.
• 58:24Also, because we identified some,
• 58:27drugs that are currently used,
• 58:29We are planning,
• 58:31target trial elimination in the
• 58:32robust research program, which include
• 58:34around ten thousand cases of
• 58:36cannabis use disorder to see
• 58:38which are the,
• 58:40outcome associated
• 58:43among the people that you're
• 58:44already using are already using
• 58:46these drugs,
• 58:47with respect to,
• 58:48and also affected by cannabis
• 58:50use disorder.
• 58:52Again, I I need to
• 58:53thank, like, my group and,
• 58:54again, for
• 58:56leading this analysis,
• 58:58the Yale Center for, the
• 59:00Science of Cannabis and Cannabinoids
• 59:01for giving us this pilot
• 59:03award, and I'm happy to
• 59:04take any question.
• 59:07Excellent work. Thank you very
• 59:08much. We have very little
• 59:10time,
• 59:11if there is any
• 59:15questions?
• 59:17No?
• 59:19Okay. Just one question I
• 59:21have. Yeah. I I've seen
• 59:23in your, initial background section,
• 59:26in your results that for
• 59:27some reason, there was a
• 59:28negative,
• 59:29association with,
• 59:31ASD spectrum.
• 59:33I mean, this is in
• 59:34the context of some studies
• 59:36suggesting that targeting the endocannabinoid
• 59:39system may may be beneficial
• 59:41for ASD.
• 59:42So it's curious that, none
• 59:44of these people are self
• 59:45medicating.
• 59:48Also, their,
• 59:49potential,
• 59:50shared,
• 59:51neuroinflammatory
• 59:52profile.
• 59:54So I wonder whether you
• 59:55have Yeah. No. That's that's
• 59:56a good point. And when
• 59:57I when I was putting
• 59:58together,
• 59:59these these slides, I was
• 01:00:01also interested about this,
• 01:00:02again, because all this more
• 01:00:04the relationship between the autism
• 01:00:06and this cannabis phenotype was
• 01:00:08different from other secondary disorders.
• 01:00:10There is also, like,
• 01:00:12one possible aspect,
• 01:00:14that, like, people with the
• 01:00:16old ishmm,
• 01:00:17they may have more risk
• 01:00:19at birth, and this could
• 01:00:20lead, like, to less exposure.
• 01:00:22But I've also, like, there
• 01:00:23could be other molecular mechanisms
• 01:00:25that are underlying this relationship.
• 01:00:26So, like, I think that
• 01:00:28there is need for more
• 01:00:29more studies about autism
• 01:00:31and cannabis to understand that
• 01:00:32there could be some treatment,
• 01:00:34or it is simply driven
• 01:00:35by population dynamics that affect,
• 01:00:37like, the participants that are
• 01:00:39enrolled
• 01:00:39and things like that.
• 01:00:41Yeah. Also because of the
• 01:00:42shared potential neuroinflammatory
• 01:00:44profile is well studied in
• 01:00:46one and is now emerging
• 01:00:47in calories.
• 01:00:50Okay. So that's my question.
• 01:00:52If there is no other
• 01:00:53question
• 01:00:56Yes. Thanks, both, Renato and,
• 01:00:59Rachel for wonderful talks. I'm
• 01:01:01I'm sure I'll reach out
• 01:01:02to you all. I have
• 01:01:03some ideas that I'd like
• 01:01:05to talk about. Thanks.
• 01:01:07Appreciate it. Thank you. Ex
• 01:01:08excellent,
• 01:01:09Tom. Thank you very much,
• 01:01:10everyone.
• 01:01:11Thank you. Bye. Have a
• 01:01:12good day.