Brain Imaging Studies with Psilocybin and MDMA - Robin Carhart-Harris

It's a real pleasure to be here

and to present the work I've been doing

over the last five years or so since my PhD.

Genuinely, I want you to try and understand

as much as you can

So it's not the easiest material,

And especially when you don't have a background in neuroscience.

It's not always that easy to get a handle on these processes.

But I'm going to do my absolute best

to try and help you understand, and

I'll provide some metaphors to try and

break things down to make things more accessible.

Also it would be useful to keep

some pens out because there will be

some references, so if you genuinely you do

want to understand how these drugs

work in the brain, then it does require

a bit of work on your parts as well, unfortunately.

So you'd have to go away and look up

some of these references and do some

background reading.

So I just want to start by saying

that this work is part of the

Beckley-Imperial psychedelic research programme,

which is an initiative between David Kotts

and Amanda Fielding of the Beckley Foundation.

Amanda is a key collaborative partner in this work,

and David Nutts, the principal investigator on it.

So we'll start with the science.

We know that psilocybin is an ingredient

in magic mushrooms.

Now, psilocybin is the [pro-drug] of psilocin,

which is remarkably similar in its molecular structure

to the endogenous neurotransmitter

which is found throughout the brain, serotonin.

So it is really quite striking how similar

it is in its molecular structure.

Just a subtle change in its structure

confers such profound effects of consciousness.

So this already

is a matter of great intrigue

about how these drugs work in the brain.

So what was found in the mid-1980s

was a strong positive correlation

between a psychedelic drug's affinity

for the serotonin 2A receptor,

a particular subtype of the serotonin receptor,

and the drug's potency.

So a good example to help illustrate

it's principle is, LSD has a very high affinity

for the serotonin 2A receptor --

it's very sticky, and it's also incredibly potent.

So that helps you understand.

Also, Franz Vollenweider did an excellent study

blocking the serotinin 2A receptor

with ketanserin, a relatively selective

serotonin 2A receptor blocker,

and he found that pre-treatment

with this drug blocked the psychedelic effects

of psilocybin. So there's good evidence

that these drugs trigger their effects

on consciousness by initial effect on

the serotonin 2A receptor.

So already we have an important

fundamental relationship that's been discovered

between the serotonin system and how these drugs work

in the brain.

So where is the serotonin 2A receptor in the brain?

Well, this is the largest serotonin 2A binding study

that's been done by a colleague of ours,

David Erritzoe.

He used a radioactive tracer, or ligand,

that sticks to serotonin 2A receptors in the brain.

And then you can detect the signal

where the ligand is stuck

And so doing this, he found that

the serotonin 2A receptor is

very much a cortical receptor.

So the outer layer of the brain, the cortex,

(it's referred to as kind of like the bark of a tree)

so this outer layer of the brain, that's where you find

the serotonin 2A receptor.

And it's especially prevalent,

especially densely expressed,

in high-level cortical regions.

So these are regions that

don't have a specific sensory function, like,

for instance, the visual cortex, which is concerned

with visual processing.

The heteromodal regions so they have a more

kind of divergent, and high-level, function

and so the serotonin 2A receptor

is especially densely expressed

in these high-level cortical regions,

such as the posterior cingulate cortex.

You will hear this term referred to again

throughout my talk

because it's a key region of the brain,

a very high-level region of the brain,

and it seems to be especially implicated in

the mechanistic action of these drugs,

how they work in the brain.

We also know that the serotonin 2A receptor

is especially densely expressed

in the particular layer of the cortex.

So the cortex is organized in a kind of laminar

way, and there's some large,

what are referred to as pyramidal neurons

in layer 5 of the cortex.

That is... (pause)

This is layer 5 here.

So there's some large neurons there,

and these are the principal output layer

of the cortex,

and there's also something else,

which is especially important about

this cellular group,

in terms of how this drug works in the brain.

So we know the serotonin 2A receptor is very important,

We know where it is,

in terms of its spatial distribution in the brain

and also within the cortex itself,

within its laminae organization. It's dense in the

deep pyramidal cells in layer 5.

We also know that if you stimulate

the serotonin 2A receptor,

you have an excitatory effect on the host cells.

So the cell that expresses the receptor,

if you stimulate it, then you're gonna make that cell

more excitable.

So these are all important principles that we know.

These are kind of the bedrock findings so far

about how psychedelics work in the brain.

But these are all quite low level;

my brain imaging work has been looking at

a higher level,

what's referred to as a macroscopic level.

So, the level you can look at and see

on a large scale in terms of

brain networks, for instance,

and regional brain activity.

So let's start with our first study,

our first fMRI study with psilocybin.

This used the modality referred to as

arterial spin labelling, which is

a method which measures changes

in blood flow in the brain.

And generally there's quite a reliable relationship

between blood flow in the brain and brain activity.

So if blood flow increases,

we generally infer

that brain activity has increased.

So this study had fifteen healthy volunteers,

mean age of 34,

the scans were 18 minutes in duration,

there was a six-minute baseline,

and then we looked at changes in blood flow

after that baseline.

There were two scans:

a placebo scan followed by a drug scan.

And volunteers just lay in the scanner.

They were presented with a fixation cross:

just a simple green cross that they looked at

on a screen, and they just relaxed

and were instructed really just to let their minds wander.

And then we looked at how

the drug affected blood flow

during these conditions.

We gave a dose of 2 mg of psilocybin--

that compares to about 50 mg when given orally.

We gave the drug intravenously,

so 2 mg is equivalent to about 50 mg orally,

so it's a moderate dose.

Here's the design. Here's our 6-minute baseline.

Infusion was given over 60 seconds.

So it's a relatively rapid infusion.

And then the onset of the effects is also rapid,

so when the drug is given intravenously,

really there is very little delay between

delivery of the drug and the onset

of the subjective effects.

So the subjective effects actually begin,

really, before the end of the 60-second adminstration,

so the drug seems to really get in the brain

very quickly and to change consciousness

profoundly very quickly.

So what was the first observation?

Well, the first thing that we get

before we analyze the results are people's

descriptions of their experiences.

So here's one of them:

This volunteer said that "there was a definite

sense of lubrication,

of freedom,

of the cogs being loosened and firing off

in all sorts of unexpected directions."

Now these subjective reports are really useful

because they give you a sense of the

mechanics that are going on in the brain,

and the changes in the mechanics,

which confer the subjective effect --

what's going on in the brain on a mechanical level

to produce the profound changes in consciousness.

This volunteer said, "Everything became

fragmented; things were all in bits

and it was very hard to hold it all together

in a coherent stream."

So it's like I said, this stuff is really useful

for understanding what is going on

on a systems-level in the brain

to produce these subjective effects.

Now, the default mode network

you've heard quite a lot about

over the last two days.

It is an incredibly important

system that's been discovered in the brain.

And one of its properties is that

it has very dense connectivity,

so if you look at the white matter

tracks in the brain,

so these are the fibers that connect

different brain regions,

then you'll find that there's a very dense

coming together of connections within

the default mode notwork.

So there seems to be

an incredibly important transit hub,

a place where different regions

can connect via,

and information can be projected from,

and also a very important integration center.

So, to integrate brain function,

information comes together at this common

convergence zone, and then

that gives a coherence to cognition, essentially.

That's how it's understood so far.

What else about the default mode network?

So here's a metaphor to help you try and understand

what people are thinking about its function.

So a metaphor that could be used to explain

what it does is a capital city in a country.

It's a place where people come together,

things come together,

business gets done.

And it's an incredibly important hub,

and if ever

God forbid, something were to happen to

London, then, the country as a whole

would be seriously effected,

and not just Britain.

So it's incredibly important

integration hub, the default mode network.

What else do we know about

the default mode network?

There's some evidences here.

The default mode network undergoes

significant ontogenetic development

from infancy to adulthood.

It undergoes maturation as cognition matures.

It has also undergone significant

evolutionary expansion

so these regions have increased

more than other regions

from primates to humans.

It's more metabolically active

than elsewhere in the brain,

so the posterior cingulate cortex,

which is the region which is circled there,

it actually accounts for 40% more

blood flow than anywhere else

in the brain.

So it is a very metabolically hungry system

and these regions that are part of it

are incredibly metabolically hungry.

It's doing something important.

Now, a matter of intrigue in neuroscience

is that people don't have a really good handle

on what the default mode is

and what it does,

but of course, they enjoy speculating,

and the researcher who really discovered

the default mode network has referred

to its very high energy levels

as being like the brain's dark energy.

So similar to dark energy in cosmology

it is something that we know is there

but we don't really know what it does.

Really, we make inferences about it,

based on its relative decrease in activity.

So when you engage in a task,

you see a decrease in activity

in the default mode network,

whereas otherwise

it is incredibly active.

So it's a mater of intrigue.

What's going on here?

What's all this energy for, and why

is it consuming so much?

We know that the default mode network

is engaged during self-reflection,

so that is a very staple finding.

We also know that during

complex mental imagery, such as

spatial navigation or imagination,

fantasy in one's mind eye,

you'll also see increased activity

in the default mode network.

Mental time travel -- so that's being

outside of the moment and

daydreaming about future events

or past biographical experiences.

So whenever you come out of the moment

and you daydream in this way,

you see increased activity and connectivity

in the default mode network.

Also, theory of mind -- which is

putting someone oneself

in somebody else's shoes.

You will also see increased activity

in the default mode network during that function.

And metacognition -- which thinking about thinking,

that's also linked with

default mode network activity.

So Raichle, who I said is the guy who

really discovered this network

has also referred to it as

the orchestrator of the self.

So all these things

led me to start thinking

having a background in psychoanalysis

and being interested in

especially Freudian metapsychology

instead of the more mechanistic

ideas of Freud. There is remarkable overlap

between his descriptions of the ego

and the relationship between the ego

and the unconscious mind, or the id,

and what we're discovering in

the default mode network.

So in this paper, with Carhart-Harris & Friston,

I submitted the idea that

the default mode network is essentially

the neural substrate source of the ego,

which is an idea to be shot down

if people find otherwise.

But that's science. That's how we work.

So what else is there about

the default mode network?

Well, what's it's relationship to depression?

There's a very interesting relationship

between default mode network parameters

and depressive symptomatology.

So we know that connectivity between

the medial pre-frontal cortex

of the default mode network

and the posterior cingulate cortex

which are

(pause)

the front bits and the back bits.

So when connectivity between these regions

is high,

then scores in patients with depression

on rumination (so these are scores in rumination.

They are thinking over problems

and ruminating on negative things),

when this is high, connectivity

betwen these regions is especially

high. And we think this system,

this overconnectivity, is really causing people to

have a kind of stereotype style of thinking.

So they are stuck in this system,

they are stuck in their own heads,

they are stuck on their sense of self.

They are usually thinking very critically

about themselves, and going over and over

about how terrible they are.

So this is a relationship we seem to be discovering

about the default mode network

and depressive symptomatology.

This provides a useful background

for what we're finding in terms of how

psilocybin is affecting

brain networks and brain systems.

So in our ASL study, we found

it was really quite surprising findings for us,

given descriptions of consciousness

being expanded by psychedelic drugs.

We are given some previous work,

for instance, by Franz Vollenweider,

we were thinking we were going to be seeing

increases in brain activity or brain blood flow

with psilocybin.

And despite dropping the threshold

and a number of different things,

we really didn't see this.

All we were really seeing was

the same pattern again and again,

which was decreased blood flow

in certain regions of the brain.

What was intriguing was that

the decreases that we were seeing were

in these very same important hub strutures

of the brain.

So, for instance, the posterior cingulate cortex,

this bit at the back,

the thalamus,

and the medial pre-frontal cortex --

so these very reliably

were coming up as being

decreased under psilocybin.

Here's just showing you again,

the default mode network is this kind of

hub, this connectivity hub, in the brain.

We also found that

there was a relationship between

the magnitude of the decrease in blood flow

and ratings of the intensity of the experience.

So, the larger the decrease in blood flow,

particularly in the anterior cingulate cortex,

the more intense people were describing

their experiences.

So whenever you find these relationships

it kind of reinforces your inferences, really,

and provides some consolidation

for what you are finding

and supports its functional meaning.

So since our ASL study

we did a [bold] study.

This is kind of the classic [signal]

of functional magnetic resonance imaging (fMRI).

We really repeated the same protocol:

15 healthy volunteers,

infused with a drug over 60 seconds at 2 mg,

and we found exactly the same thing.

So this was really nice reinforcement

for the initial finding that we had found

with ASL.

And these regions where there was a common

overlap between the decreases in blood flow

with ASL and the decreases in,

essentially venous oxygenation,

or oxygenated blood with the BOLD signal

of fMRI.

So one of the merits of BOLD fMRI is

that it allows you to do these

functional connectivity analyses.

So just to give you a feel for what that is,

here is an image which shows

the default mode network in orange

and let's concentrate on the default

mode network for a moment.

So you can see that there are two regions in it,

and there is one that has has yellow text,

the PCC? and then you can see there is this

time series underneath.

So the PCC time series is in yellow

and you can see that it overlaps with

anothertime series, and that is

the medial prefontal cortex.

So it is by looking at correlations between

fluctuations in the BOLD signal

that we identify functionally coherent

brain networks. We know these regions

work together as a common network

doing a common function.