- Welcome back, everyone. For those of you that have
been in breakout rooms, this is our penultimate session and I'm going to ask
you to take your seats and I will introduce Dr.
Mignot who is in the building. Okay. Elvis is in the building. I feel like my job is complete. So Emmanuel Mignot then. He is the Craig Reynolds
Professor of Sleep Medicine in the Department of Psychiatry
and Behavioral Sciences at Stanford University and he's the director of the
Stanford Center for Narcolepsy. Dr. Migno
t is recognized as
having discovered the cause of narcolepsy. Born in Paris, he received his MD and PhD there
and then practiced medicine and psychiatry also in
France for several years before serving as a visiting scholar at the Stanford Sleep Disorders
Clinic and Research Center. He then joined as faculty and director of the Center
of Narcolepsy in 1993 and he was named Professor
of Psychiatry in 2001. Dr. Mignot received has
received numerous awards for his work, including and most recently
t
he 2023 Breakthrough Prize in Life Sciences. He is a member of both
the National Academies of Sciences and Medicine. Welcome, Dr. Mignot. (audience applauding) - First, I want to thank
you so much for being here and for listening to my talk
and thank you for inviting me. Now I'm going to ask you to
do the introduction every time 'cause it's so much better
with an English accent, don't you think? (Emmanuel and audience laughing) - [Claire] I did cut yours down. I think I had three pages printed o
n you, but I actually almost
just said here's Emmanuel. But anyway.
(Claire laughing) - Anyway, so today I'm
going to talk to you about idiopathic
hypersomnia and its cousins. The reason I mentioned its cousins because I think this concept of idiopathic hypersomnia is changing and you know, what is hypersomnia
is not completely clear. So the first question I will go through before discussing what should
be done in the future is what is really idiopathic hypersomnia. So I guess this will work. We
'll see. Okay, so I have many conflict of interest, but because I work with everyone
that's trying to, you know, develop drugs for
narcolepsy and hypersomnia. So the good thing is since
I have conflict of interest with everybody, I'm less conflicted. But still you have to know that. So the first question I
wanted to go through is what is idiopathic hypersomnia? Where does it come from? Because I think often, you know, we forget about our history,
but history is very important. It's really our ro
ots and
often it gives us a little bit of a better idea of what
people were thinking initially. So, you know, the term idiopathic
hypersomnia really started in the Czech Republic, when
it was still communist, by Bederich Roth, who, by the way, was
a very incredible man, which I never met, unfortunately. But he passed away. He resisted the occupation
by the communist regime and he had a lot of
difficulties because of that. And he really defined
idiopathic hypersomnia, as I will explain, with
a se
ries of symptoms that corresponded to something
different from narcolepsy 'cause, you know, you had
narcolepsy with cataplexy and REM sleep abnormality and the idea was maybe
idiopathic hypersomnia would be something more that
has to do with non-REM sleep. You know, being more like
needing more non-REM sleep as opposed to REM sleep for narcolepsy. Then the 1980s, you know, Dr Honda, which was in Japan, also tried to differentiate different
types of hypersomnia and for him, he called
monosymptoma
tic needs. There were narcolepsy with cataplexy and then monosymptomatic narcolepsy and then essential hypersomnia. And then in the 1990s, the
MSLT started to be developed to diagnose narcolepsy type
one, narcolepsy with cataplexy. It was really designed
to diagnose patients with narcolepsy type one. So people who have narcolepsy cataplexy to really verify that
they had to diagnoses in an objective way. But then it started to be
applied to everybody else that complained of sleepiness and then ju
st started to
define these disorders based on the result of the test. Narcolepsy type two when
you're positive for the MSLT but you don't have cataplexy
and idiopathic hypersomnia when you fall asleep quickly but you don't have cataplexy and you don't go into REM sleep
and really where are we now. And then in parallel with this, there was this odd condition
called periodic hypersomnia or Klein-Levin syndrome
was first described in 1925 by Klein and then Levin, which I will talk a little bit abou
t it, which is unlike the other
condition was something that fluctuated so people
would be very tired and very sleepy and
needing a lot of sleep, a little bit like idiopathic hypersomnia but more extreme but only
during periods of time and then get completely cured and then it would restart by episodes. So I will discuss a little bit about all these different conditions. So again, the first definition of
idiopathic hypersomnia come from Bederich Roth. We really had this idea that,
again, people
had already, oh. It's is moving spontaneously. Had this idea that that
you had narcolepsy, people had discovered in the 50 that narcolepsy were going
very quickly into REM sleep. So there was these ideas that indeed you have narcolepsy type one, which is a REM sleep problem, you know, like people have this
hallucinations, this dreaming. And that was one type
of problem with sleep. And then people had discovered
they were non-REM sleep, which is like stage three, four, or two. And then the idea i
s idiopathic
hypersomnia where people, instead of needing more REM sleep, needed more non-REM sleep. And the idea is that it would
be really difficult to awaken, be in sleep, you know, very deep sleep, a little bit like a child, you know, when they have a lot of slow wave sleep and would sleep long period of time. And that they had difficulty waking up. It was much more like people
were kind of sleep deprived and needed more sleep but felt like they were
always sleep deprived. And then it differ
entiated, you know, a little bit this monosymptomatic
idiopathic hypersomnia where people will sleep a
long time during the day and the night. And then some other polysymptomatic
idiopathic hypersomnia where they were also the same
plus were a little confused and difficult to wake up, you know, having like a lot of sleep inertia. And then he mentioned this, which is not very fashionable
idea, of neurotic hypersomnia, which was was more like hypersomnia due to a psychiatric disorder, which is bas
ically people who, you know, would be tired all the time but when you look at their sleep, it didn't seem to be very abnormal. And he mentioned also that
it was frequently familial, that maybe there were genetic component, and really the MSLT was
not considered important at the time. Nobody was doing the MSLT but they were doing the sleep studies and what they wanted to see
in the sleep studies is people who had a lot of slow
wave sleep, you know, needed a lot of sleep
that's not REM sleep. And
a little bit later, you know, Dr. Billard in France in
Montpellier really started to discuss the fact that sometime it's a little bit difficult to differentiate idiopathic hypersomnia from some form of depression
that needs a lot of sleep. And I think, you know, it's a little bit academic, it's always a problem because, you know, depression of psychiatric disorders, they also come from the brain. It's not like, you know, you can separate, you know, psychiatric and neurologic. You know, the only
reason psychiatric and neurology has been separated is because of a couple people in the 1800s at the end of the 19th century. One guy that was called Babinski that kind of was doing
all these exams in people and then trying to see if one of the nerves were not working. And if he found an objective finding, he would say, "Ah, this is
a neurological disorder." And if he didn't find
anything, he said, "Ah, this is a psychiatric disorder." But the truth is of the
matter is the more you look, the mo
re you might be
able to find abnormalities and then things will disappear. If we apply this definition, maybe one day we'll be able
to analyze the entire brain and then nothing else will be psychiatric because, of course, everything
originate in the brain. So in Japan, they had a
slightly different view. The Japanese were very
focused on cataplexy so they definitely defined this narcolepsy with cataplexy very well. In fact, there was this
professor called Honda who was insisting that cataplexy w
as the most specific symptom, was the most important for narcolepsy, which is still true to date. But then he was saying that
there were some other patients that looked like narcolepsy in the sense that in narcoleptics, it's not like they sleep too much. It's more like they cannot
stay awake so they will wake up and then they will have to nap regularly but their nap would feel refreshing, which is not the case at all
in idiopathic hypersomnia. The classical picture of
idiopathic hypersomnia, peo
ple cannot nap, they
have trouble waking up. Again, they have a lot
of this non-REM sleep and you know during non-REM
sleep you have sleep walking, which is really confusional arousal. So it's more like people
who have trouble being awake and in non-REM sleep. And he was trying, he said that there were
probably a gradient of people who had more like narcolepsy type one but with refreshing naps and were still a little bit
the same as a narcoleptic. And then there were some people who were more on
the
idiopathic hypersomnia, more non-REM sleep. But you see, the idea at the time was to differentiate REM sleep hypersomnia and non REM-sleep hypersomnia. That was the whole idea. And non-REM sleep hypersomnia would be idiopathic hypersomnia and REM sleep hypersomnia
would be narcolepsy. The problem is that then
people started to say, "Oh, if that's the case..." By the way, interestingly, which is something that
I want to point out 'cause this is coming now. We know that when you have
narcolep
sy with cataplexy, you usually never improve. I mean, if you don't get treated, you have the disorder for life
and here it just follow up. Our people felt like 10 years after, you know, having developed a disorder and he showed that only
like 10%, you know, felt better when they
had narcolepsy type one. So narcolepsy type one is
definitely a permanent disease, whereas for idiopathic, what they call essential
hypersomnia syndrome, so this mix between
narcolepsy without cataplexy and idiopathic hy
persomnia, you know, about half would improve. So again, that's something
that I always mention to people with idiopathic hypersomnia, even so I really want to
treat you very actively, I just want to make sure that
you know that there are cases that get better with time and I think that's very
important for children or young adults so that you
don't put them a sticker that is going to live all their life because maybe at some point, they may be able to stop
medication and feel better. So in addi
tion to this concept, there were also this concept
of recurrent hypersomnia that came from Klein
and Levin with this idea that you had these people
who were completely normal and then suddenly, they
will become extremely sleepy for two or three weeks and they will sleep enormously
like 20 hours or 24 hours. And even when they were
awake, they were not normal. They were totally confused. It's almost like they could
never completely wake up. They were in a daze, you know? And even sometimes their
brain was disconnected so that they would be illogical or maybe sometime be
disinhibited, for example, eat some very strange food
in a very primitive way or even have sexual disinhibition. You know, like, really like
the cortex was disconnected. They were almost like half
in a state of half dream, which they say frequently that they say that they
feel they're in the bubble, that the world is kind
of feeling different. But what was very unusual
is that this would go on for two or three weeks and
then boom, they will
be completely normal after and then it would relapse
from time to time, very unpredictably. So they call this periodic
or recurrent hypersomnia and we call this Klein-Levin syndrome. And it's definitely a special entity because, you know, these people have this, they're usually adolescent men and then usually, when you
wait, it gets better with time. So after 10 years or 12 years, often the episode become less intense and then people usually get
cured of their hypersomnia an
d it's responsive to lithium, which suggests that maybe some overlap with some of this psychiatric diagnosis. So again, besides this classification
that was more clinical, what happened is the MSLT. So the MSLT was invented
really to diagnose narcolepsy, cataplexy, narcolepsy type one
and it's a good test for it. It's 95% positive in narcolepsy and 95% negative in controls. So it's a very good test. It has only four or 5% false
positive and false negative. So it's really a good test
for narcolep
sy with cataplexy. So problem is we created
a condition based, instead of basing it on
the symptom of patients, we based it on the test. So we knew that narcolepsy
cataplexy was good at the diagnosing type one narcolepsy and we say, "Oh, now
we are going to use it for everybody else that's tired
and if they have REM sleep, we'll call them narcolepsy type two and then if they don't have REM sleep, we will call them idiopathic hypersomnia." And because it will be non-REM versus REM. So problem is
nobody had done his homework and in fact, it means nothing. Basically, you can have an
MSLT with 2SOREMP one day in a patient with narcolepsy
without cataplexy. If you repeat it three months later, it looks like idiopathic hypersomnia. Or you get idiopathic hypersomnia, you repeat it three months later, it becomes narcolepsy without cataplexy. So practically now we really
realize, probably too late, that this test doesn't mean anything. You know, it's much more important to focus on the symptoms
of
the patients than on the test. So for me the test, the only interest of the
test is the insurance. You know, unfortunately, insurance don't always consider
narcolepsy without cataplexy or idiopathic hypersomnia the same way. Some of these insurance will
pay better for medication when you have an MSLT with multiple SOREMP than when you have MSLT without SOREMP. But truly and honestly
it's completely idiotic. It doesn't mean anything. So you just have to convince... (audience applauding) You j
ust have to convince the insurance because the MSLT is not a good test to really predict what
people really experience when they have idiopathic hypersomnia or narcolepsy without cataplexy. So I'm just hoping that this artificial thing will disappear but we don't know when it will happen. So the first kind of stroke against the MSLT was
published quite a while ago, you see, in 1996, where
people trying to see if people have an MSLT with SOREMP or an MSLT without SOREMP, does it correlate with pe
ople having more difficulty waking up or having naps that are not refreshing? Does it correlate with this
idea of non-REM hypersomnia and REM hypersomnia of Bederich Roth and Dr. Honda? And what they found is absolutely not. They found that, first, there
were a lot of false positives. They were false positive that we knew that about 5% of cases
had a positive MSLT, but sometime it was just by chance. And we found that also in
the general population, if you take people out of the street and you g
et them an MSLT, about 4% are going to have SOREMP and look like a narcolepsy type one. So 4% is not a big deal when
you have a real narcoleptic that has cataplexy and everything because it's 96% of
the time it's positive. So that's okay and it's a good test
but then when you apply to the general population, it's not good. And one of the things that
was found is the SOREMP during the MSLT are
confounded by a few things. So for example, if you have shift work, the circadian clock regulate
REM sle
ep very strongly. So in the morning hours, you have a lot of REM sleep naturally. So if you come from another country and you do the MSLT during the day, it's very likely that you
have will have REM sleep because your circadian clock
will be in the wrong time zone. So that was a big, one of the big confounder. We found that a lot of normal people, if they're shift worker they
get SOREMP during the MSLT and then some others get it by luck. I mean, we don't know because
if you repeat it again, it
doesn't change anything. It's also a little bit
related to sleep deprivation. If people are chronically sleep deprived, they have more tendency of having SOREMP. But in general, there's definitely a group of false positive people
who are kind of normal and have a positive MSLT even without complaining of anything. And finally, I mean, as I mentioned, people realize more recently
that the MSLT with SOREMPs or narcolepsy without cataplexy is really not a stable phenotype. If you take people who ha
ve
a positive MSLT and a SOREMP and sleepiness and no cataplexy and if you repeat it,
only about 10% repeats. So most of the cases, they switched to idiopathic
hypersomnia or vice versa. So again, idiopathic hypersomnia and
narcolepsy without cataplexy, I'm repeating myself, it
shouldn't be differentiated. We just don't have any evidence that it's different based on the MSLT. Now that doesn't mean that
there is not different types of narcolepsy without cataplexy
and idiopathic hypersomnia, but i
t's just that the
MSLT is the wrong way to differentiate people who are tired. You know, we have people
who are very tired and some report that they
have long sleep episode. Some people, they feel
refreshed after an nap. Maybe that's much more
important than the MSLT. That's what I wanted to say. So in conclusion, narcolepsy without cataplexy
is not a reliable diagnosis and this has been confirmed by many people based on the MSLT. So one of the first
question you may ask me, which is totally rea
sonable, is now we know that idiopathic hypersomnia and narcolepsy without cataplexy should not be differentiated. So does it matter? I mean, maybe we don't care. Maybe they react to the same treatment. And indeed, there is
a strong point with it because, for example,
if you use Xyrem now, we have used a lot of Xyrem in patients with idiopathic hypersomnia and
narcolepsy without cataplexy and it seems that it
helps a subset of patients whether or not you have
idiopathic hypersomnia or narcolepsy
without cataplexy. I haven't found a difference. So definitely the MSLT
shouldn't be the predictor for trying to select the drugs. That's what I want to point out. And then, you know, probably boost narcolepsy
without cataplexy and idiopathic hypersomnia, they might be a circadian phase problem with this REM sleep occurrence. We don't know. It hasn't been studied. And then in some cases, definitely there is some
overlap with psychiatric issues. But the problem with psychiatric
issues are so com
mon. You know, like 15% of the
population has anxiety or depression. So you know, what does it mean? You find it often associated
and it overlaps with it, but, you know, maybe it's
connected, maybe it's not. But certainly, that's one important aspect. And again, I mentioned that
for idiopathic hypersomnia and narcolepsy type two, we don't know the long term prognosis. There have been very few studies that have looked at
people who were diagnosed and then 10 years later,
are they off medication,
are they doing well, or do
they still have the problem? And as I mentioned, people
who have looked at it, it seems that probably about
half improve with time. So it's very important to remember that. So now we killed the MSLT. So the next question is... (audience laughing) It is very nice, but what do we do then? You know, whoop. So what do we do? So one of the thing is, you know, some people try to cluster, you know, you look at patients and
you say do patient cluster in certain groups? And unf
ortunately, some
people try to cluster but with the MSLT and they
kind of got some symptom and PhD clusters. But honestly, I don't
know what you all think, but I didn't think that
was very convincing. They found like seven clusters, but I mean, it was a bit difficult to really make sense of what they found. And I think the main problem was really because they were adding
the MSLT into the mix. I think we really need to cluster more by not with the sleep MSLT, but maybe more with nocturnal sleep
or with sleep during the day. And indeed even for nocturnal sleep, people haven't found big differences between narcolepsy without cataplexy and idiopathic hypersomnia. So I think we need some new
methods to differentiate because now we have idiopathic hypersomnia and narcolepsy without cataplexy. And by the way, it's really annoying
because in all my write-ups, I always write narcolepsy
without cataplexy slash idiopathic hypersomnia. And this way I try to get the
reimbursement of narcolepsy wit
hout cataplexy and then, you know, so that I can get whatever drugs. And even if it's MSLT, I don't care. And then if someone complains,
I haven't been caught yet. I will say, look at all this paper. It means nothing. So that's why I call them the same way, but hopefully, I will win
that battle if it happen. But the insurance company,
thankfully, their IQ is very low, so they're very... (audience laughing) So they don't. That's the only place where I
think artificial intelligence should quickly
replace them. But I think definitely it's going to be, we have to redefine it and we have to redefine it in
some things that matters more. So we want to really see
what people experience during the day because night, okay, it's important but
that's necessarily the problem. The problem with idiopathic hypersomnia and narcolepsy without
cataplexy is during the day. People feel they're foggy or people feel they
have to take these naps. They may feel better after naps. They feel worse after a nap. T
his is a real phenotype
of the patients that we see and that's what we have to treat. And that's why I think we have to switch from studying night sleep and the MSLT to studying during the work day. And I think there's
definitely new devices now that are being developed where you can do EG during the entire day. And this was is done in
actually Europe, they do that. They measure sleep and wake
with EG all during the day. But the problem is
often they force people, like in Montpellier, they force
people to lay down in a bed and do nothing and they just
see if the people fall asleep. I don't think it's a natural setup. And so one in Italy,
they do slightly better. They try to use the PSGs,
you know, measures of sleep, but people can walk around
or do whatever they want and then they try to see what happened. I think the future, ideally, would be if you could wear, like, an helmet and then you could walk
around, do whatever you want, cook, et cetera and we will record your
sleep and your
EG pattern during the entire day and entire night. So we'll see, oh, do
people go into micro sleep? Do they have periods of
time where they have, they looks like sleep but they're awake? Or do people take very long naps or even it's possible
that some of these people who are report brain fog, a lot of patients report,
like, being not there. You know, if you look at the EG, it might be abnormal but
without really sleeping. But nobody has really studied that. So I think a lot of people now realize
that what we need is shift the focus from looking at sleep at night to looking at wake during the day. Can we find something
that looks like foggy, you know, being foggy? Can we look at something like fighting your sleep, et cetera. Do you go into REM sleep? And it's clear that
that's where I am going and, like, everybody else is going. But in addition, this is okay because you can maybe describe
how people feel better and try better accuracy of the same tone in terms of the recording. But we a
lso want to go to
the cause of the problem. And for that, I think
doing some biological tests like genetic and proteomics
could be very helpful. In particular, I believe that a lot of
patients that are tired, sometimes we don't know
if they're are tired because their circadian clock is abnormal. Some of these patients with
idiopathic hypersomnia, I'm sure some of you, you have a lot of difficulties
waking up in the morning. That's one of the classic
idiopathic hypersomnia. But one of the problem
s, that this problem is also found in people who have delayed sleep phase syndrome. Like, you all know that adolescence, you know, when you're younger
and you go to bed at two, you can't go to bed before 2:00 AM and then you can't wake up at 7:00 AM. You're just like a zombie. So it looks a little bit
like idiopathic hypersomnia for that symptom. And imagine that maybe some people with idiopathic hypersomnia, maybe it's one of their problem
that their circadian clock is completely abnormal
so th
ey try to wake up in the middle of their real night and that's why they feel so bad. And in addition, they
sleep during the day. So if the cause of this
patients is partially circadian, we should definitely try
to treat them differently from other people in
particularly with light or even with this new melatonin compounds that are very powerful to maybe
reset the circadian clock. So that's why the other
thing that I feel we have to do is to use some more
biologically based, you know, test to try
to understand
a little bit better different subtype of
idiopathic hypersomnia that could be easier due
to not having enough sleep or having the wrong circadian phase, being living in Tokyo time or a combination of both or not having enough wake promotion. Do we need to help people
to sleep better with Xyrem or do we need to wake
them up with stimulants? So these kind of things
could really help us to better, you know, classify
and treat these patients. And one that's, you know, so there's a lot
of progress
in artificial intelligence that are making the way we
record sleep much more efficient. For example, this new program that was
developed by Matthias, who is a very bright computer scientist, can score sleep within like half a second. So it can tell you, I'm simplifying, but it can tell you you are in REM sleep during that half second, you know, for very high resolution instead
of using every 30 seconds. So for example, now we can try to see if there is increase like REM sleep event
in the
middle of the day for one second and maybe that's much more predictive of narcolepsy type one or certain type of narcolepsy type two or certain type of idiopathic hypersomnia. And we are looking at this
kind of things during the day, but this is really preliminary, but what I'm doing now is I'm, you all have been through this MSLT, so I'm trying to use the data. Unfortunately, I would like to have people wear these helmets and have sleep during
the day and the night for hundreds of people
with idiopathic hypersomnia and analyze that, but I don't have it. So what do we do? We use what we have, which
is the MSLT, and the MSLT, normally you take these naps, but in between the nap, they ask people to try to stay awake. So we are extracting, you know, these periods of
wakefulness between the naps and then we're trying to
see if it looks different between different subtypes
of idiopathic hypersomnia and narcolepsy without cataplexy. So right now, we started by
just looking at micro sl
eep, you know, like, do people
kind of fall asleep for a very brief period of time? Like, you know, nod off
in between these naps where they're supposed to stay awake. And indeed, we're finding
that some people do and also people don't. I think that will be a
much more objective way of separating different
types of hypersomnia. But this is just the beginning. So I think again, I think our next step for this area is to define what is being awake. You know, that's really key for me. Like, people w
ith idiopathic hypersomnia, they tell you they're tired but it's very different
from one person the next. Some people they can't resist sleep. They fall asleep. Others they tell you they're
just, like, never really awake, but it's not like they fall asleep. So we really need to
differentiate maybe these patterns and maybe find some
things that predict this. So the ideal case would
be to have patients with idiopathic hypersomnia
report, "Oh, I'm very foggy", or "Oh, I did fall asleep, et cetera",
and correlate what people report with EG and then maybe find some ways
of predicting this objectively with EGs then we'll
have a biological basis for the problem that could really be used as a diagnostic test. And I'm hoping that this will be the kind of things we need to do. But unfortunately, we
don't have the data yet. And I think that's the biggest problem. What we need is to really
have a lot of people wear EG during the daytime and annotate how they
feel during the daytime. And that's the
future in my opinion. But unfortunately, it's
going to take a few years to be able to have that kind of data. The second thing is maybe
we don't need the EG. Well, after all, like using the EG, it's very uncomfortable. Even so, as I told you, it's really becoming much better now. There's this kind of
helmets that you can wear that you could wear for
24 hours or 48 hours, but it's still, maybe it's not necessary and you all know what an actigraph is. You know, for example, the iPhone, the Apple
watch, they have an actigraph. There is Fitbit. All those, they all have actigraph. Basically it is to measure
movements that are high frequency and maybe it would be enough, you know, we don't know. We just have to test with actigraphy. Can we differentiate people
with idiopathic hypersomnia in subgroups? So I think this still need to be done. And for example, even
for narcolepsy type one, people have done some work that seems to show that even with low
resolution, like 30 second epoch, which i
s not very good, now we can measure activity every second. It can differentiate narcolepsy type one from other hypersomnia. But I think we need to do a lot more with better devices and
idiopathic hypersomnia narcolepsy without cataplexy to try
to find different patterns. You know? And with that, I mean, in my opinion, I'm always coming back to genetics because genetics is a cause of problems and it's often, you know, a completely different
way to look at things. And what's very clear from
all th
ese experiments is when you really only use sleep or only use one modality or actigraphy, sometimes these programs they work but they always make mistakes because you can foresee everything. For example, we have very good program now that can look at sleep at night and can predict narcolepsy
as good as the MSLT, like 95%, et cetera. But if the person doesn't
sleep at all, what do you do? So that's a problem. You may still have narcolepsy. You can't observe anything. And it's very complicated. Th
ere's always exceptions. There's always cases that don't fit. So that makes it a problem because if you want to
screen for narcolepsy in the general population, for example, you really want to make sure that you have 100% specificities, that it can pick up even rare cases. And that's why it looks
more and more apparent than to have the best performance
for all this program is to mix up different types of data. So one will be sleep or actigraphy, which will measure certain things but add the gene
tic or add
proteomics, as I will argue. And the combination of
these different markers that are totally different will make much more better prediction because it doesn't make
the same type of mistakes, you know, and that's why it's so powerful. So genetics is one of the way, and by the way, we know that people, that's one of the surprise
that has been found that people who report sleeping too much or needing too much sleep say
often is a slight correlation with the genetic risk
factors of bipol
ar disorder and schizophrenia. We don't really know why, but there's a little bit of overlap between the genes that
predispose with too long sleep and the genes are predisposed
to this kind of problem. And, of course, if we could
have like a genome studies of idiopathic hypersomnia, maybe
we'll find genes that will, you know, as a total can predict and help to predict these problems. And one thing that I feel very strongly about is finally trying to use proteomics. So maybe that's a little
bit t
echnical for you, but the genetic is you're born with it, it predicts things but
never very strongly. You know, it's not like 100%. You know, we are all born with genes but it depends what we
do with it, you know? I mean, depending of our
environment, et cetera. A lot of people are born
predisposed to narcolepsy type one, but only a very small
percent develop narcolepsy because it depends of
what kind of flu you got, what kind of things you experience
during your entire life. And I think it's th
e
same for most disorders. So genetics will never
tell you, oh, that's it. You have narcolepsy or you
have idiopathic hypersomnia. It could tell you you are more at risk of, and I think because it's a
completely different type of information, it will complete and improve prediction when you use that plus actigraphy and plus what people experience. And proteomics is a little
bit closer to physiology 'cause the gene produce proteins, but the proteins also
change with the environment. So for exampl
e, you know circadian rhythm, we can now predict circadian rhythm by measuring all these proteins and some protein rises at
different times of the day and then if they rise at
the wrong time of the day, you can tell if the circadian
rhythm is completely abnormal. So that's what we are doing now. We're measuring like 5,000
different proteins in the blood, which in one little drop of blood, and then some of these proteins, they increase in the early morning, others increase in the middle morning,
others increase around lunchtime, some other increase in
the middle of the night. And then by looking at all this
protein and where they are, we can kind of have an idea
if we take a blood sample and we note it's 8:00 PM
and then we test it and say, oh no, not at all. It says it's 12:00 PM. So in that case, you know, we can know that the
circadian phase is abnormal of four hours, you know, because the protein profile tell us. And we are also trying to
find a group of proteins that reflects the f
act
that people sleep too much or groups of proteins that
people are sleep deprived. And my dream, you know, would be to use that plus
the genetic plus something that measures sleep during the day. Then we'll know if
people have micro sleep, how is their sleep during the day, how is their sleep during the night. We'll know what are the genetic factors that predispose idiopathic hypersomnia, different types, how
is the protein changing so that we will know if these
people live in Tokyo time or in
, I don't know why I use Tokyo. I should use Paris time. Paris time or in Tokyo
time or in the normal time, we could see if the protein
profile kind of suggests that they don't have enough sleep. In that case, we will
give probably more Xyrem. Or not enough wake, and in that case, maybe you
want to give an orexin agonist or a stimulant. And I think that will really
help to treat patients better. I'm going to pass this, I think. It's just a general idea. And just to mention where maybe we have be
en
moving a little bit more, of course, there is narcolepsy but also Klein-Levin syndrome. We did quite a bit of work
on Klein-Levin syndrome. I mean, I think I need now
to work a little bit more on idiopathic hypersomnia. But the reason I work on
periodic hypersomnia is that it has been defined
very well since the 1920s. Nobody contests the disease. You know, nobody fights. Oh, it's MSLT, not the MSLT. It has a clear definition and we know it's a very severe disorder. Like, people suddenly, for
two weeks, they look like idiopathic
hypersomnia multiplied by 100. You know, they're just
like completely in a fog. They can't even talk. They just are very abnormal. It's very, very strong phenotype and it's very odd because
it goes on and off. And when you do even like
scanning of the brain activity during hypersomnia, when
they are very tired, their brain cortex seems
to be disconnected, which may explain why they have a lot of problems responding to questions, et cetera because it's so str
ong that some parts of the brain
don't seem to work normally and you can detect it by doing scans of looking at the activity of the brain when you have an episode or not. And then we did a genetic
study and it took a long time, there's a lot of photos, as you see, to take blood sample from tons. You all know that I'm
an international vampire and I always get, you know,
blood sample from anyone, but it's for a good cause
and it's to do the genetics and these proteomics and for example, when we di
d a genetic scan of the people with Klein-Levin syndrome, we found that one of the top gene that was found was also
involved in bipolar disorder, which is interesting because
it also react to lithium. So I suspect that there
is probably some forms of hypersomnia that overlap with this kind of pathology. And by the way, there were
not only this strong gene, but what was very interesting is that it also repeated very well. So if you take one patient, you know one gene is good, but what you want to
also do is a combination of genetic factors. So for example, when we took the first 500 patients with Klein-Levin syndrome, we could calculate like a polygenic score, we could find all the genes
that are associated with it, but maybe not at a very high level. And when we tested a series
of 200 patients after that, it was also quite predictive. So it repeated. So the idea here is not
to look for one gene, but maybe a combination of
many genes that can predict, you know, the risk of the disease.
And the proteomics. The same way. By measuring 5,000 proteins, we could actually diagnose patients with KLS relatively well. I mean, we need to
definitely reproduce that, but I was really surprised myself. It looks like in the blood, even when people are not in episodes, we could find a signature
of like 20 or 30 proteins that were elevated or
decreased in patients with Klein-Levin that seems
to predict relatively well is that Klein-Levin and not. And it was reproduced and we
found it both in th
e blood and the CSF, which comes from the brain, which makes me think that
it's probably a true finding. So now I really want to do
that in idiopathic hypersomnia and narcolepsy without cataplexy. 'Cause my dream would be to really be able to combine all these different modality. You know, again, my point is that if we use only
sleep, we'll miss something. But if we use sleep and
genetic and proteomics, I think at the end, we'll
become much more specific and be able to really understand
differen
t subtype of people who have sleep problems. So in my opinion, what needs to be done in
IH is we definitely need to study more sleep and wake. And for this we need to use, you know, deep learning and new statistical modern to predict brain fog,
to predict macro sleep, to predict different types of wakefulness and different types of sleep. And I think we need to
do it at home, you know? Like, people need to wear
something at home for 24, 48 hours and then we can really see
what's happening object
ively during the day. Are they sleeping, are they not sleeping, et cetera. I think we need to definitely look at the genetic idiopathic hypersomnia. It's really too bad
because it hasn't been done and it'll be easy because
there's so many patients. But even just the GWAS, I think, would be very informative. And then I'm big believer
in this proteomics. That by measuring these proteins, we can find if people have
an abnormal circadian clock, if people have sleep deprived
so they need to sleep mor
e, or if they have insufficient sleep or wake promoting mechanism. And I think that's another promising way to separate different
types of hypersomnia. And I hope to be doing that in the future. And really definitely at the end maybe will
differentiate different subtype of idiopathic hypersomnia and
narcolepsy without cataplexy. They won't be based on the MSLT. I'm sure. In my opinion, there would be
based much more on symptoms, you know, like people will just sleep
a lot even during the day, pe
ople who have brain fog but don't sleep, people who have micro sleep. But we'll see. And then, of course,
what will be important is treatment response because we don't care if they all work, the same drug all work, but I suspect that if we define biologically
different disorder, we'll see much more clearly. Like, for Xyrem, for example, I mean, some of you idiopathic
hypersomnia, it's amazing. I mean, some of these patient
reacts super well to Xyrem, they do very well, but for
others it doesn't
work at all. And I personally haven't found a clue why Xyrem works in some
idiopathic hypersomnia and not others. And maybe by doing this subtyping, we might be able to better
tailor the type of treatment to each type of idiopathic hypersomnia. So that's my dream. And I just want to point out
that this is the main people who have been working with me in the lab. Katie Cederberg, Germain
Kolosov, and Adrien Specht. For notably in proteomics,
Aditya Ambati, for the genetics. And a lot of the work
about
this circadian proteomics is done in collaboration with Harvard, in particularly Jeanne Duffy and Chuck. And for narcolepsy and all these samples, often I work with the entire world and I just want to mention Isabelle Arnuff who works a lot with KLS. And Giuseppe Plazzi and
Yves Dauvilliers works a lot on narcolepsy. And finally, I want to mention I work also with people in Takeda
because they are, you know, head with this orexin
agonist and Dmitri Wolson and other, we are now are good col
leagues and we try to, they have
more resources than me. So I guess I work with whoever has, and I think they're very interested in detecting narcolepsy
and some of these disorders for their drugs. So I have some collaboration
going on with them. So thank you so much for your attention. (audience applauding) - [Claire] Thank you very much. We actually have time for one question and while the KLS the panel
come and get mic'd up. So one question only, I'm afraid. Who's gonna have it? - One questio
n so make sure it's the right question.
- [Claire] Okay, your hand was first. - That's a lot of power. - [Claire] Better be good. - [Audience Member] I
think it's a question that everybody's got on their minds. Oh, you're talking about
a complete paradigm shift. You're talking about
taking the present system, tearing it down, and
building something else up. What else? Or excuse me, how long is this gonna take for all of us? - Yeah, it's going to take a while. Unfortunately, you know, is what act
ually with the
epistemological obstacle. You know, it's very difficult to change some things that are ingrained. It really takes sometimes a generation. I don't think it will take a generation because everything is accelerated, but I think it will take five years if we are very optimistic, probably 10 years until we are there. I wish it would go faster,
but it's very difficult. There is like investment, you know, like all these people
with money, with the MSLT. There is no reimbursement. There i
s a drug, the insurance companies
that want to make sure that they continue to
build this and not this. They don't want to pay. So unfortunately, logic is
not always a top priority. We all know that, but it
will eventually prevail. We just have to push. And your voice, by the way, is
one of the way you can push, you know? You have to really make your voice heard and say that this situation has to change. And in particular, I really believe that
idiopathic hypersomnia has to be considered real di
sease. Not, you know, like an afterthought. Oh, you don't have
narcolepsy without cataplexy. Yep, you have idiopathic hypersomnia. - [Claire] Yeah, am I right in thinking that you are actually not going to retire? (audience laughing) - Well, I don't know. It depends. - [Claire] Okay, I think we do have time for one more quick question
while they get ready. Andrew. - [Andrew] I'd like to
just follow up actually on what you were saying, Claire, and what you were saying, Dr. Mignot. What can we in
this room do to shorten that timeline? Should we, for instance,
all be doing 23andMe? Does that help you with
any of your research? Should we be wearing actigraph wristbands so that we can get you more data? How can we help you? - Huh, that's a good question. Certainly I think participating in research is definitely
one of the best way, whether it's my research or
the research of someone else and I think that really
makes a big difference. I would say that it's also important to be a tiny bit po
litically active in terms of making sure that these disorders
are not forgotten by NIH and other people who try to fund research and then even lobbying for, you know, like the SM and some of those, you know, conventions so that they
change the system faster. I mean, for me, you know, like, I call that death by committee. I mean, as soon as you put
like 20 people together, they start to argue forever. And then at the beginning, they say, we are going to change everything. And then at the end, oh
yes, now we are going to change the way we define sleep latency. Instead of taking like
two epochs of non-REM and whatever, it will be
only the first epoch of N2. Victory. And then just like you
have spent six months discussing together and you end up with a very mediocre result. I'm not sure how we
can change that because that's inherent to some
of the group thinking. But I think, for example, in the context of idiopathic hypersomnia narcolepsy without cataplexy, I think there is a good
impetus
to change that. But we also are very afraid
because you know, Karen and I, I'm sure we agree, but then if we change it and
then the insurance company to say, oh, that's great. Now we don't even pay for
narcolepsy without cataplexy, you know? So I can't really, you can do the right thing scientifically, but be shoot in the foot by, you know, by the insurance company. So, unfortunately, the best you can do, I think at least I cannot advise you what to do for this bigger area. So for sure, working
on the
ICX 64 will be very important. But I think the research at the end will train off. If there is enough data that shows that this is exactly the
way that it should be done, it will work. I'm not really answering your question. I wish I could, but if you want to give your
blood samples, that's okay. - [Andrew] Well, I'm sure
we've got people waiting at the door ready to take everyone's samples. - But I think it's a very, we'll make more, and also I think another thing that's definitely unde
restimated is I believe in single individual
making a big difference. And I think we don't have
enough people and, you know, that are trained and
interested in this area. I think, you know, having younger students and
people really have a future in this area of idiopathic hypersomnia and narcolepsy without
cataplexy is critical. I really feel we don't have
enough people doing this kind of research and, you know, funding fellowship or things like that I think can make a big difference. - [Claire]
Thank you very much. Okay. (audience applauding)
Comments