"Exploring the Relationship between Sleep Apnea and Stroke" Jacqueline Geer (09/07/2022)

September 23, 2022

ID8111

To CiteDCA Citation Guide

00:00So now I will turn the session over to
00:02Doctor Yaggi, who is our program director
00:04for Yale Center for Sleep Medicine,
00:06and he is going to introduce Doctor Dear.
00:10Thank you, doctor Hilbert.
00:11It is my distinct pleasure to
00:14introduce Doctor Jackie Gear,
00:17both as a new faculty member in our section,
00:20having just finished her fellowship
00:22training and as our inaugural
00:25speaker this academic year and
00:28our Yale Sleep Medicine Seminar.
00:30Dr Gear completed her undergraduate
00:32training at Washington University
00:34in Saint Louis Medical School, NYU.
00:36And then we were fortunate to
00:38recruit her to Yale, where she.
00:40Really has stayed for her clinical
00:42training both as an intern,
00:43a resident, a chief resident,
00:46and fellowship training in pulmonary
00:48critical care and Sleep Medicine.
00:50And she has been incredibly productive
00:53during her fellowship and has managed
00:56to advance in a very short time.
00:58Our understanding of sleep disordered
01:01breathing and intracranial hemorrhage,
01:04which is a very common type of stroke,
01:07results in considerable mortality as well as.
01:11Functional and and cognitive
01:13impairment and morbidity.
01:15She has already received numerous awards
01:17and grants for her work in this domain,
01:20including salary support from the NIH,
01:22NINDS and stroke net during her
01:24fellowship training a TS funding
01:27through the Aspire Fellowship,
01:29which is designed to promote
01:31sleep research among pulmonary and
01:33critical care medicine fellows.
01:35She's also received an American Academy
01:38of Sleep Medicine physician Scientist
01:40Training award and most recently.
01:42The Yale.
01:44But why CI RCTs,
01:47a clinical scholars KL2 award all
01:50related to her work at this interface of
01:53intracranial hemorrhage and and sleep apnea.
01:56And today she's going to talk to us
01:58more broadly about sleep apnea and
01:59stroke and share with her some of the
02:02work that she has has done in this domain.
02:04Welcome,
02:05Jackie.
02:08Wow. Thank you so much car for
02:10that warm introduction and thanks
02:12for everyone for being here today.
02:14It's an honor to give the first
02:16sleep seminar of the semester.
02:18I actually hadn't realized that
02:19initially when I signed up for the state.
02:21So lots of lots to live up to you
02:24or or maybe not so much we'll save.
02:27So I am extremely excited today to talk
02:30to you all about sleep apnea and stroke.
02:32It's a topic of great academic
02:36interest to me and.
02:38Before we get started, let me just.
02:44There we go, and before we get started,
02:46I'll just put the CME slide up again.
02:48I think you all already saw it,
02:50and doctor Hilbert went through the
02:53the rules and I have no personal and
02:56relevant financial disclosures to share.
03:00So before I delve into
03:02the outline for the talk,
03:04I'll just take a minute to share with you
03:06how I became interested in this domain.
03:09And so as a first year pulmonary
03:11and critical care fellow,
03:13I had an early rotation in the Neuro
03:17ICU where I had many stroke patients,
03:19but also many ICA or interest
03:22cerebral hemorrhage patients,
03:23which is not something I had seen
03:25much of as a medicine resident.
03:27And then there was a a great component.
03:30Happenstance as I think is
03:32typical for people's research on
03:35trajectory and career stories.
03:37So it was,
03:38first of all the the time of year in
03:41that mid fall where you're supposed
03:45to figure out what you're going
03:47to do with your copious research
03:48time in the second and third years.
03:51And and so I was sort of pondering that
03:54and happened to be working with them.
03:57Kevin, Chef in the neuro ICU,
03:59was my attending.
04:00Time and Kevin is not only
04:02a world class researcher,
04:04but also an outstanding clinician and
04:06has a a strong mentorship track record.
04:10And so he and Clara had actually
04:12just received R1 funding for a
04:14sleep and a schematic stroke trial,
04:17which I'll come back to later on in the talk.
04:19But Kevin wondered if I might have
04:21any interest in becoming a stroke
04:23researcher and he took me out for
04:25Arethusa ice cream to discuss it more.
04:27And how could I say no?
04:29And so I.
04:30It really was interested in the topic,
04:33which I learned a lot about the relationship
04:36between sleep apnea and ischemic stroke,
04:39but wondered if maybe the
04:40same was true for ICH,
04:41which I think was the big question.
04:43I discovered that there was a
04:45huge gap in the literature and
04:46there's actually hardly any work
04:48that had been done in that domain.
04:51So I was intrigued.
04:52I had a dream team for my from
04:55my mentors and so we just started
04:58working and have been continuing it.
05:01Ever since.
05:01And so.
05:02Today I'm gonna talk to you about the
05:06evidence that links sleep apnea and
05:08strokes are sort of what do we know,
05:12where are we?
05:13Explore the proposed mechanisms.
05:15So why, why is it,
05:18you know,
05:18why is sleep apnea a risk factor for stroke?
05:21And then talk to you at the end
05:22about the role of treatment.
05:27So first we'll review the evidence.
05:30And so I know I don't have to go into too
05:33much detail for this particular audience,
05:35but since it is the beginning of the
05:36academic year and we have learners with
05:38variable experience, I'll just take
05:39a moment to talk about sleep apnea.
05:41And so OSA is a common form of sleep
05:44disordered breathing and it results
05:46from intermittent upper airway collapse
05:48leading to apneas and hypopneas,
05:49which causes cycle of abnormal Physiology,
05:53which includes hypoxia,
05:55sympathetic activation,
05:57hemodynamic disturbance and
05:58disruption of sleep architecture.
06:01Sleep apnea effects up to 40% of the
06:04general population in the literature,
06:06and probably more.
06:09And it's an established an independent
06:11risk factor for ischemic stroke,
06:12which is present in up to 75%
06:16of this patient population. Um.
06:20So some background on stroke.
06:22So it's the third leading cause of death
06:25in the United States and the 2nd globally.
06:27And there are almost a million strokes a
06:30year in the United States and more than
06:33150,000 stroke related deaths every year.
06:36And so scheming stroke is the
06:38most common stroke subtype,
06:40followed by interest,
06:41cerebral hemorrhage ICH.
06:44Stroke is the leading cause
06:46of long term disability,
06:47and it costs an estimated $35
06:50billion a year related to healthcare,
06:54medication and missed days of productivity.
06:59So if you were specifically
07:01on spontaneous ICH.
07:04Which accounts for a substantial
07:05portion of all strokes,
07:06but I think we are less familiar
07:09with so a common research
07:11definition of ICH is non traumatic,
07:14abrupt onset of severe headache,
07:16altered level of consciousness,
07:18and or focal neurologic deficit,
07:20with an associated focal collection
07:22of blood within the brain frankoma
07:24seen on neuroimaging or autopsy.
07:26It accounts for up to 20% of all strokes
07:30and has a 50% thirty day mortality,
07:32and the the half who do survive tend to
07:37have significant functional limitations.
07:40And risk factors include hypertension,
07:42age, diabetes, smoking,
07:44anticoagulation,
07:45particularly warfarin.
07:46And so you'll you'll kind of note
07:48that these are are fairly similar
07:50risk factors to ischemic stroke.
07:56So there have been a number of studies
07:58providing epidemiologic evidence
08:00and solidifying the connection
08:01between sleep apnea and stroke,
08:03and I I just have a sampling of
08:05them listed here on this slide,
08:07and I'll draw your attention
08:09to the rightmost column,
08:10which shows a consistently strong
08:14association of high prevalence of OSA
08:18in the respective stroke populations
08:21relative to the general population.
08:24And actually our own doctor Moseman
08:26was one of the first to make
08:28this observation and you can see
08:30his case control study in 1995
08:32at the top of the list there.
08:34And there have been tons
08:36of studies to follow.
08:38And as you can sort of see from
08:40the study population column,
08:42most of them are looking at
08:44ischemic stroke and Tia and a
08:46few of them are indiscriminate.
08:48And at the bottom,
08:49you can see a reference for a study
08:52that I conducted during fellowship where
08:54we look specifically at interest rate,
08:55real hemorrhage patients and and
08:58matched controls and found that
09:01there was also a very strong
09:03prevalence of OSA in the cases,
09:05which was significantly more
09:07than in the controls.
09:08And I'll tell you more about that later.
09:11So the real question arising from many
09:14of these early studies was whether sleep
09:16apnea was a risk factor for stroke,
09:19or if stroke caused sleep apnea.
09:25So the directionality of association
09:27was unclear based on the studies
09:29I just mentioned, and ultimately
09:32it appears to be bidirectional,
09:34but strongly favoring sleep
09:36apnea as a cause for stroke.
09:39So some evidence to support the the
09:42notion that stroke can cause sleep apnea
09:46can be found in some earlier studies
09:49looking at neuromuscular weakness
09:51of the upper airway dilator muscles
09:54which can lead to airway collapse.
09:57And individuals who are
10:00presumably predisposed.
10:02Far more compelling is the notion that
10:05sleep apnea is a risk factor for stroke.
10:09And support for this comes from many studies,
10:11and the reasoning is is broad,
10:16including the fact that sleep apnea
10:19is strongly associated with Tia,
10:21in which no residual deficits remain.
10:23And studies have shown that sleep apnea
10:27patients with vulvar neuromuscular
10:29weakness and continue to have sleep apnea
10:32even after recovery from their deficits.
10:35Additionally,
10:35we see that the presence of OSA is generally
10:40indiscriminate of location or stroke subtype.
10:45Which suggests that stroke
10:47is less likely the cause.
10:48And then finally sleep apnea is
10:51strongly associated with wake up stroke.
10:54So we'll delve more into these
10:56mechanisms shortly.
10:57And there's also the issue of
11:00confounding by hypertension,
11:02diabetes, hyperlipidemia and obesity,
11:05and which I will also try to
11:08address in a bit.
11:12So as many of you are aware,
11:15our own Clara Yagis Fellowship project was
11:18actually a groundbreaking contribution
11:20to the domain of sleep apnea and stroke.
11:24And was one of the first to
11:27prospectively tackle this chicken and
11:29egg on sort of causation question?
11:32So in his observational cohort study,
11:35consecutive patients
11:36underwent polysomnography.
11:38And subsequent strokes and deaths
11:41were verified and recorded.
11:44And the diagnosis of OSA,
11:46I should mention,
11:47was based on HIV 5 or greater,
11:50and those without OSA served
11:52as the comparators.
11:53So what they found was that 68% of
11:57enrolled patients had sleep apnea,
11:59and that the presence of sleep apnea
12:01was associated with significantly
12:03higher likelihood of stroke or death,
12:06even after they adjusted for confounders.
12:10So this is depicted in these
12:13two items from their paper here.
12:15So on the left you can see
12:17the Kaplan Meier curve.
12:18And and they also found that the
12:20severity of sleep apnea correlated
12:22with the likelihood of stroke or death,
12:25which you can see on the in
12:27the table on the right.
12:29You can see that the severity of
12:32sleep apnea is stratified with an
12:36increasing likelihood of association.
12:41And so this table from the same study
12:44shows the unadjusted and adjusted
12:47hazard ratios for the risk of
12:50stroke or death for all covariates.
12:52And in the unadjusted analysis,
12:54traditional cardiovascular
12:55risk factors that we think of,
12:58like age, diabetes,
12:59and hypertension were all associated with
13:01an increased risk of stroke or death.
13:04But what I'd like to draw your attention
13:06to is actually the bottom of the
13:08table where sleep apnea is listed,
13:09and the findings here show.
13:11Basically a doubling of risk.
13:16Of of stroke or death from any cause.
13:20And so this remains significant
13:21even after adjustment for all
13:23of the relevant covariates,
13:25which supports the hypothesis that sleep
13:28apnea independently increases the risk
13:31for stroke and all cause mortality.
13:33So this study was the really
13:36the first of its kind,
13:38but many subsequent studies have
13:40reproduced these results and and then
13:43gone on to also explore Physiology
13:45mechanisms and things like that,
13:47which we'll discuss shortly.
13:52Um, so under the mentorship of Clara,
13:55Yagi, and Kevin Chef.
13:56As I mentioned earlier,
13:58I studied the relationship
13:59between sleep apnea and NIH,
14:01specifically during fellowship.
14:04And I was really shocked to learn
14:06that so little research has
14:08been done in this sub domain.
14:10I still am sort of surprised but
14:12but it's it's a lucky break for me.
14:15And so utilizing a large case
14:18control data set that included
14:203000 acute IH patients and 3000
14:23matched controls called the
14:25called Eric or the the Eric trial,
14:28all of all of the patients in both arms
14:32like 9098% of them had completed a Berlin.
14:35Questionnaire at the time of
14:37study enrollment,
14:38so we performed a cross-sectional analysis.
14:42And given that patients were enrolled within.
14:46Hours of their acute IH event we
14:49we made the leap that their ruling
14:52questionnaire should be reflective
14:54of their pre stroke status since not
14:56much time had passed and what we
14:59found is highlighted on this slide.
15:01So basically you can see that obstructive
15:04sleep apnea was significantly
15:06more common in those with ICH as
15:09compared to matched controls,
15:11and that this finding held even
15:12after we adjusted for significant
15:14covariates and universal.
15:16Founders.
15:18We also did some special analysis because
15:21as many of you are probably thinking,
15:24we too were thinking that,
15:26you know, hypertension is a really
15:29important risk factor for IH,
15:31potentially more so even than than
15:34schimick stroke and it's also,
15:37you know, associated with sleep apnea.
15:39So we,
15:40we perform some additional analysis
15:43to kind of investigate this further.
15:45So first of all.
15:48There I should mention that there are
15:51sort of two bucket locations for IH.
15:53There's non low bar,
15:55which is usually due to hypertension
15:57and then there's low bar which is more
15:59commonly due to cerebral amyloid angiopathy.
16:02And so we stratified our results
16:04by location and found that this
16:07that that our initial finding
16:10held regardless of location,
16:11IE even for low bar strokes
16:14which were presumably not due to
16:17hypertension as the primary driver.
16:19And we also did a hypertension focused
16:21analysis which supported our findings.
16:27So now that we've gone through
16:29a review of sort of evidence
16:31linking sleep apnea and stroke,
16:33I'd like to spend the the next several
16:36minutes talking about proposed mechanisms
16:39for how sleep apnea causes stroke.
16:44And I realize this is a very busy
16:47slide with a lot of words and I'm
16:50not going to go through them all.
16:52So that's a sort of more
16:55to to show you all this,
16:57this web of interconnection
17:00among risk factors and outcomes.
17:04And I will distill it down in the
17:08end to to six mechanisms that I'd
17:11like to discuss today, but I.
17:14I just want to sort of highlight
17:16that the physiologic sick quality
17:18of of obstructive sleep apnea can
17:21lead to these chronic downstream
17:24mechanisms of atherosclerosis,
17:26including inflammation,
17:28hypertension,
17:29impaired glucose metabolism
17:32and snoring related vibration.
17:35And then there are additional
17:37physiologic stresses related to sleep
17:40apnea which can also predispose
17:42to stroke and sort of potentiate.
17:44A lot of these other mechanisms,
17:46and these include.
17:49Acute decreases in cerebral blood flow,
17:52a heightened risk of Afib,
17:55induced cardio embolic stroke,
17:58paradoxical embolism through PFOA,
18:00which happens to be much more
18:03common in in people with sleep apnea
18:05than than the general population.
18:07And so we'll spend the next
18:09several minutes kind of delving
18:10into these mechanisms.
18:15So the six mechanisms I'd
18:18like to highlight for today's
18:20talk are intermittent hypoxia,
18:22sympathetic activation,
18:24metabolic dysfunction, snoring and
18:27vibration induced carotid disease,
18:30arrhythmia and cardioembolic stroke,
18:32and PFOA and paradoxical embolism.
18:34And all of these mechanisms have
18:37been extensively studied and and
18:40have been shown to have a direct
18:42causal effect on the pathway
18:43between sleep apnea and stroke.
18:49So the first mechanism
18:51is intermittent hypoxia.
18:53So sleep apnea results in
18:56repetitive episodes of hypoxia
18:58and the varying severity and
19:00duration all throughout the night.
19:03This intermittent hypoxia and has
19:06been shown to lead to generation
19:10of reactive oxygen species,
19:12activation of transcription
19:14factor and release of cytokine
19:17and adhesion molecules.
19:19And this ultimately results in
19:21endothelial dysfunction and the
19:23development of atherosclerosis.
19:27So as our barzman and colleagues
19:29conducted an interesting and relevant
19:32study a few years ago looking at
19:35outcome implications of the hypoxic
19:37burden and they were able to
19:39actually quantify the hypoxic burden.
19:41And so, given that that HIV poorly
19:45predicts outcomes in sleep apnea,
19:48the authors thought that maybe
19:50quantification of the actual burden of
19:53hypoxia and including sort of depth and
19:56duration metrics would more accurately
19:59correlate with adverse cardiovascular
20:01and cerebrovascular outcomes.
20:04So they use two very large cohorts
20:07of older adults.
20:08They use the outcomes of sleep
20:10disorders in older men cohort and
20:13the sleep heart health cohort.
20:15And they measured hypoxic burden as the
20:18respiratory event associated associated
20:21area under the desaturation curve.
20:25And they compared this curve
20:27with the pre event baseline.
20:29So in the figure from this paper on
20:31the left you can see an example of
20:34this calculation with Apneas recorded
20:36on the top and oxygen saturation on
20:39the bottom with time on the X axis.
20:42And the figure on the right is
20:45from the same study and it shows
20:49adjusted survival for cardiovascular
20:51more mortality across.
20:53Each severity category of hypoxic burden.
20:59And they did a nice accurate rainbow
21:04for for different severities.
21:06So you can you can see here with the X
21:09axis showing time and years and the Y
21:12axis I'm showing survival probability
21:14that the worse the hypoxic burden,
21:17the shorter the survival.
21:24The next mechanism by which
21:26sleep apnea increases stroke
21:27risk is sympathetic activation,
21:29and this figure on seeing here on
21:32this slide is a nice depiction of a
21:36representative cut from nocturnal monitoring.
21:39But you can see that there's a spike in
21:43sympathetic nerve activity with each apnea,
21:46and a corresponding surge in blood
21:48pressure at the termination of each apnea.
21:51And so this finding is reproducible
21:54basically with each apnea.
21:56And so it's happening presumably
21:59hundreds of times per night and more
22:03with increasing severity of sleep apnea.
22:06And so this is thought at least in
22:08part to be due to a resetting of
22:11baroreceptor sensitivity during sleep
22:13in patients who have sleep apnea.
22:16And this finding is most pronounced
22:19in end to sleep. Random sleep.
22:24Interestingly, this phenomenon of
22:27of increased sympathetic activation
22:30can also be seen during wakefulness.
22:34And so in an older but excellent
22:36Physiology study by summer and colleagues,
22:39the authors demonstrated that this
22:42sympathetic nerve activity actually
22:44persists during wakefulness in
22:46people with obstructive sleep apnea.
22:49And so here you can see on nerve
22:52tone tracings in awake subjects.
22:55Normal, meaning no sleep apnea
22:58and OSA in patients with sleep
23:01apnea and so on the left,
23:03you can see that there's lower
23:05lower tone than on the right,
23:07and this is again an awake patient.
23:10And so very clearly,
23:12this propensity towards heightened
23:15sympathetic activity is actually
23:17not just limited to when when these
23:21sleep apnea patients are sleeping,
23:23but also throughout the day,
23:25which increases risk further.
23:32So moving on to mechanism 3,
23:35I will talk about metabolic
23:37dysfunction and I'd specifically
23:39like to focus on type 2 diabetes
23:42in patients with sleep apnea.
23:45So this is relevant because
23:47diabetes on its own is actually
23:51a potent risk factor for stroke,
23:54but OSA is also an independent
23:57risk factor for diabetes,
23:59even after adjustment for confounders
24:02like BMI and weight gain.
24:07So this table here on the right is
24:09from an observational cohort study
24:12where over 1200 veterans in the
24:15VA system were evaluated for sleep
24:17disordered breathing and surprisingly
24:20about half did not have diabetes at
24:23the time of the of study enrollment.
24:26And so these folks went on to
24:28complete a full PSG and the study
24:31population was divided into quartiles
24:34based on sleep apnea severity and.
24:37The main outcome of interest
24:40was incident diabetes.
24:41So their initial analysis looking at
24:44linear Trend showed a significantly
24:47increased risk of incident diabetes as
24:50a function of severity of sleep apnea.
24:53And as you can see here in the table,
24:55sleep apnea was associated with diabetes,
24:58even after adjustment for age, race,
25:02gender, and baseline fasting glucose, BMI,
25:06as well as weight gain or change in BMI.
25:13And mechanisms through which sleep
25:14apnea is implicated in the development
25:17of type 2 diabetes include activation
25:19of the sympathetic nervous system,
25:22which we just discussed because sympathetic
25:25nervous system activity plays a key role
25:28in regulation of glucose metabolism.
25:30Another mechanism is oxidative stress,
25:33where sleep apnea may accelerate
25:36progression to type 2 diabetes.
25:391/3 mechanism is subclinical
25:42inflammation and adipocyte derived
25:45inflammatory factors like I6,
25:48which have been well established
25:50mediators in the pathogenesis of diabetes.
25:54And systemic inflammation
25:55and cytokine production,
25:57which are augmented by sleep
26:00loss or sleep deficit,
26:01can also hasten the progression to diabetes.
26:04And then finally,
26:05it it seems in in more recent
26:08literature that OSA in pregnancy has
26:10actually been associated with insulin
26:13resistance and gestational diabetes,
26:15and this is an area of active research.
26:22So the 4th mechanism is one that I I think
26:27is kind of fun because it's so intuitive.
26:30So snoring and vibration
26:33induced carotid disease.
26:35So basically the direct effect
26:39of of snores on the local carotid
26:43artery causing disease.
26:45So the effect is not limited
26:47just to people with sleep apnea,
26:50because anyone who snores is at risk,
26:53but since almost everyone
26:55with sleep apnea snores,
26:56it is irrelevant risk factor.
26:59So snores originate in the upper airway.
27:02They cause vibration of the
27:04pharyngeal wall and local structures
27:06which are all near the carotid.
27:09And then over time when the
27:12carotid endothelium continuously
27:13is exposed to this vibration,
27:15it can become damaged and atherosclerotic.
27:19The evidence to support this
27:20comes from a lot of studies,
27:22but one particular study of
27:25interest by Lee and colleagues is
27:28noteworthy because they compared
27:30carotid and femoral arteries and
27:32snores and found that the carotids,
27:35which are directly affected by snoring
27:38vibration had higher atherosclerosis
27:40burden than the femoral arteries.
27:43And as you can see here on the slide,
27:45the Audrey ratio was was 10 even
27:47after adjustment for relevant.
27:49Variables.
27:51It's also worth mentioning
27:53that in the SAVE trial,
27:56they did evaluate self reported
27:58snoring patterns and stroke events in
28:01high risk patients with sleep apnea,
28:03and the authors observed that
28:05among the sleep apnea patients,
28:07increased frequency,
28:09self reported frequency,
28:10and loudness of snoring was
28:13associated with greater risk
28:14of cerebrovascular events.
28:22So. Mechanism 5,
28:26arrhythmia and cardioembolic stroke on.
28:30So we often jump to Afib as the
28:32the the main arrhythmia that we
28:35associate with sleep apnea and stroke.
28:38But sleep apnea is in fact linked
28:41with higher risk of many arrhythmias.
28:44So Mara and colleagues compared
28:46the prevalence of arrhythmias in
28:48two samples of participants from
28:50the sleep Heart health study,
28:52matched for age, sex, BMI,
28:57and race and ethnicity.
28:59And they found that among
29:01those with sleep apnea,
29:04which for the purposes of the
29:06study was defined as an RDI of
29:09greater than or equal to 30,
29:11and that there was a significantly
29:14higher rate of arrhythmia.
29:17And and not just A-fib.
29:20So in the figure on the left,
29:21you can see various arrhythmias
29:23listed on the X axis,
29:26with the Gray bars representing the
29:28those with sleep apnea and the white
29:32bars on patients without sleep apnea.
29:35And you can nicely see that all
29:38arrhythmias noted have vastly higher
29:41prevalence in the sleep apnea cohort.
29:44And on the right,
29:46you can see a table from the same
29:49study and looking at adjusted and
29:52unadjusted odds ratios really
29:54relating arrhythmia with sleep apnea.
29:57Um,
29:58so?
29:58All arrhythmias are obviously harbingers
30:01of poor cardiovascular implications,
30:03and so this is relevant outside of Afib,
30:06but Afib in particular has a another hit on.
30:11And because it can give rise
30:13to cardio embolic stroke.
30:14And so we've also noted in in
30:18various studies when it comes to
30:20Afib that there does seem to be
30:24a dose response relationship and
30:26where by increasing sleep apnea
30:29severity is associated with higher
30:32prevalence of Ethan and and also
30:34we've seen that in people with
30:38paroxysmal atrial fibrillation.
30:41Um,
30:41these paroxysms tend to emerge
30:43shortly after an apnea or hypopnea.
30:51So I'm moving on to mechanism 6,
30:55PFOA, and paradoxical embolism.
30:57And PFOA is a risk factor,
31:01a well known risk factor
31:02for cryptogenic stroke.
31:03It's more prevalent among patients with
31:07sleep apnea compared to those without.
31:10And sort of mechanistically speaking,
31:13frequent transient right side pressure
31:16increases during the apneas themselves
31:19can cause right to left shunting through
31:21the PTFO which can lead to stroke.
31:24And then also may contribute to nocturnal
31:27desaturation in the setting of shunting.
31:35So now that we've established that
31:37OSA is a risk factor for stroke and
31:40explored many of the key mechanisms,
31:43I'd like to focus the remainder of the
31:45talk and hopefully some discussion about
31:49the the on the treatment implications
31:52and future directions in this domain.
31:59So why is this important?
32:00Well, currently acute
32:04ischemic stroke treatments,
32:06which have gotten quite excellent,
32:09are not widely applicable because of narrow,
32:12narrow therapeutic windows
32:14or accessibility issues.
32:16And there are basically no
32:19targeted therapies for for
32:20hemorrhagic stroke or Tia.
32:26So a logical hypothesis then is that PAP
32:29therapy will improve stroke outcomes
32:32and may reduce recurrent events.
32:34And it makes sense because, you know,
32:37we just talked about all of these
32:40mechanisms which are exacerbated by.
32:42Apnea burden.
32:43So if we reduce the apnea burden,
32:46then perhaps we can improve outcomes.
32:49And if this is true,
32:50then pat therapy and stroke patients
32:53may actually represent a new frontier
32:56in treatment which we really need.
33:02So first of all,
33:03what do we know about PAP and outcomes?
33:07And unfortunately, really not too much.
33:11You know, I even even went back to the
33:13literature as recently as yesterday
33:15and there's really not much there,
33:17and that's not for lack of effort.
33:19And we basically know that PAP
33:22use among post stroke patients
33:24appears to be associated with
33:26a decrease in stroke mortality.
33:29But this is largely based on observational
33:32studies and examining treatment efficacy,
33:35which are really subject to
33:37healthy user effect and thus
33:40tend to overestimate the effect.
33:43And for those unfamiliar with
33:44the healthy user effect,
33:46this is a phenomenon where patients
33:48who are adherent to one therapy,
33:51so in this case PAP,
33:53may also be adherent to other therapies
33:56or may value a healthier life.
33:58Style in other ways and other
34:01domains which may lead to an
34:04overestimate of the therapy of choice,
34:07of the effect that the therapy of choice.
34:11So we also know that RCT is
34:14to date have been small,
34:16so typically under 80 patients and
34:18have shown mixed results and the mixed
34:21results are thought to be multifactorial.
34:24But some issues or ideas as to why
34:27they're so mixed include variability
34:30and path initiation timing.
34:33Small sample sizes as well as varied
34:37endpoints such as you know, intermediate
34:41outcomes of cardiovascular disease,
34:43functional recovery, mortality.
34:49So what are the guidelines say?
34:51Well, before 2021,
34:53they didn't say much of anything.
34:56And now, as of the 2021
35:00update from the American Heart
35:03Association and Stroke guidelines,
35:05there's actually an entire section
35:07dedicated to obstructive sleep apnea,
35:09which is exciting and I think important.
35:12And so the recommendations are distilled
35:14down to the figure I'm showing here,
35:17but basically.
35:18They're saying that in patients with an
35:22escape chemic stroke or Tia and sleep apnea,
35:25treatment with positive airway pressure can
35:28be beneficial for improved sleep apnea,
35:30blood pressure,
35:31sleepiness and other apnea related outcomes,
35:34which is sort of a bucket of
35:38things that I think a lot of
35:40investigators are trying to teeth out.
35:42And and this is based on moderate
35:45evidence per their classification
35:48system and they also recommend that
35:51in patients with an ischemic stroke
35:54or Tia and evaluation for OSA may be
35:57considered to diagnose sleep apnea.
35:59And this is based on weak evidence
36:02by their classification system.
36:05And so I think, you know,
36:07the takeaways from here are that
36:10first of all.
36:12You know I think a lot of us believe
36:14that we probably should be testing all
36:17stroke and Tia patients for sleep apnea.
36:20But we can't really provide the
36:22evidence to back it up because of all
36:24of the reasons that we just discussed
36:26and all the issues with them with how
36:29do you study this population and and
36:31and keep it consistent so that you can
36:34actually make a a statement that is
36:37of better than weak recommendation.
36:42You'll also note that there is
36:44no mention of hemorrhagic stroke,
36:47and that's because of the large gap in
36:49literature which I mentioned earlier.
36:55So I'd like to take a few minutes to
36:57highlight some of the ongoing trials
36:59here at Yale and and around the country.
37:02And the first is rise up,
37:05which is basically looking at the
37:08question of whether PAP improves
37:11post stroke outcomes and what the
37:14ideal timing of PAP initiation is.
37:17And I'll also mention that my
37:19Co mentors Kevin, Jeff and Char
37:21Yaggi are Co Pi on this project.
37:24And so the study is actively enrolling
37:27patients with acute ischemic stroke in
37:30the hospital and performing ambulatory
37:32polysomnography in the acute setting.
37:35And just to mention,
37:37this technology is actually groundbreaking
37:40in and of itself to be able to do
37:43basically a full PSG in the hospital,
37:46you know,
37:47the day or two after an acute stroke
37:50definitely has the potential to build
37:54new diagnostic pathways for stroke
37:57patients if the outcomes warranted.
38:00So the study aims to determine if CPAP
38:02helps after stroke and if there's a
38:05difference in the degree of improvement
38:07between timing of acute intervention,
38:10which would be one week following stroke
38:14or sub acute intervention at one month.
38:17And the main outcome is functional
38:20recovery at six months following stroke.
38:24So the study was initially designed to
38:26enroll patients only with ischemic stroke,
38:29but we have since expanded it and to
38:33enroll an observational arm of patients
38:36with interest cerebral hemorrhage and
38:39given the strong association that we
38:42saw in my previously discussed work and
38:45the infrastructure through rise up and
38:47the mentorship team to be able to do it.
38:54Another study that I would like
38:56to mention is sleep smart.
38:58So this is a national sleep
39:00apnea and stroke treatment trial
39:02and Yale is a clinical site,
39:04and the aim here is to test whether
39:07a treatment of sleep apnea with pack
39:11therapy after a recent stroke or a
39:14high risk Tia prevents a recurrence of
39:17stroke or ACS or all cause mortality.
39:22As you can see from the the flow
39:25sheet protocol here and patients are
39:28screened and then again ambulatory full
39:32PSG is performed on the first night.
39:35Some patients don't qualify because
39:37they don't have sleep apnea,
39:38they don't have obstructive sleep apnea
39:40or they're limited to central sleep apnea.
39:43But if patients have an HIV of
39:4710 or greater then they are they
39:49they move on to a run in night.
39:52Time basically a trial of CPAP
39:54for the second night,
39:56and then some are not able to tolerate it,
39:59in which case they are not randomized.
40:01But those who are able to tolerate
40:04CPAP for four or more hours are
40:07then randomized into either the
40:09intervention group or the control group.
40:17And then finally,
40:18I'll just mention my own pilot study,
40:20which I alluded to briefly when
40:22I was telling you about rise up.
40:24So the evaluating neurological
40:26recovery in NIH and sleep apnea
40:29study or the enriched SA study,
40:32because everything needs a
40:34snazzy name to get funded,
40:36is a pilot project of mine that's
40:39currently funded by the ASM
40:42Foundation for one year this year.
40:45And I plan to enroll 20 patients
40:48with intracerebral hemorrhage in
40:50the hospital setting and perform
40:52that same wonderful ambulatory
40:54polysomnography that I've been telling
40:57you about in the acute setting.
41:00And my aims are to determine the
41:02prevalence of sleep apnea and it's
41:05individual physiologic traits and ICH.
41:08And to compare functional outcomes,
41:12which is the primary outcome,
41:14as well as cognitive status,
41:16which is the secondary outcome at three
41:19months post stroke among patients with
41:22among IH patients with and without OSA.
41:26And my hypothesis is that those
41:28with ICH who have sleep apnea will
41:32have worse outcomes at three months
41:34compared to those without OSA.
41:37And that measures of more severe sleep
41:40apnea will correlate with worse outcomes.
41:44And so I'm just,
41:46you know the sort of the,
41:48I think the most exciting aspect of this
41:51to me is actually doing the objective
41:54gold standard testing because while
41:57my previous work was very helpful
41:59to sort of highlight an association
42:02and one of our limitations was that
42:05the Berlin questionnaire was our
42:07our sort of diagnostic tool because
42:10there were no polysomnograms and so,
42:12so it's very exciting to be able to.
42:15Actually perform this.
42:17And so here you can see on the the plan.
42:23So so the patients will have a
42:28bunch of baseline assessments
42:31including certain metrics for IH,
42:35predicted outcomes and mortality as
42:40well as cognitive testing and then other
42:44and then other testing and then they will.
42:47Follow up at one month and then at three
42:51months and in this particular study
42:54there testing will end at three months.
42:56But as Clara mentioned at the beginning,
42:59I was just recently funded for KL2
43:05award in which we plan to look a lot
43:08more at the the phenotypes and and
43:12Physiology questions and actually study
43:14these patients out to six months time.
43:17To see their outcomes.
43:24Time. So on this slide,
43:26I've highlighted some of the knowledge gaps,
43:29many of which we kind of
43:31discussed through the talk,
43:33but just to lay them out there.
43:36They include who will benefit from CPAP?
43:39Because we don't want to force anyone
43:41to use CPAP if there's no reason to,
43:43because patients typically don't like it.
43:46And what is the ideal timing of
43:48testing and treatment initiation
43:50in relation to stroke onset?
43:52What dose and what type of PAP is most
43:55likely to work and to promote adherence?
43:58And what is the role of phenotyping?
44:00And is all that holds true for
44:03ischemic stroke and Tia true for ICH?
44:06And on a personal level,
44:08I am trying to tackle a few of these.
44:11And you know in in particular,
44:14I think that that final question
44:16and while it seems like it will
44:18be true and it makes physiologic
44:20sense that it would be true,
44:22it'll be really nice to actually show it.
44:24And so, so time,
44:26time will tell.
44:31So in summary, sleep apnea is independently
44:36associated with Tia and stroke,
44:40both ischemic and ICH,
44:43and the association appears somewhat
44:46bidirectional but vastly favors the
44:50direction of sleep apnea causing strokes.
44:53And mechanisms include intermittent hypoxia,
44:57sympathetic activation,
44:58metabolic dysfunction,
45:00shortened sleep duration,
45:02snoring and vibration
45:04induced carotid disease,
45:05arrhythmia, and PFOA.
45:09Diabetes, hypertension and Afib
45:10appear to be on the causal pathway
45:13between sleep apnea and stroke.
45:17Challenges exist and conducting long term
45:20randomized controlled trials with CPAP and
45:23the benefit is not yet clearly defined.
45:26And then finally, I hope I've impressed
45:28upon you that there are a lot of new
45:30and exciting stroke trials underway.
45:31So I think that hopefully if
45:34I'm invited back to speak,
45:36you know in in several months to
45:40years that we will have more to say
45:42on these topics and maybe more some
45:45more concrete or definitive answers.
45:48So with that, I want to just thank
45:50everyone for your attention today.
45:51It really is a huge honor to talk about
45:54this topic with this particular group.
45:56And I definitely look forward to hearing
45:58questions as well as any thoughts and
46:01suggestions that anyone has as I start
46:03my own research career here at Yale.
46:06And before I take questions,
46:08I just want to thank my research mentors,
46:10specifically Clara and Kevin.
46:12I know it sounds corny,
46:14but they really did believe in me
46:16and my potential and when I had
46:19very little research background.
46:21So I'm forever grateful for that.
46:24And of course,
46:25here's an obligatory picture of
46:26my 3 little people who keep me
46:29very active and sleep deprived.
46:31And thank you again,
46:33and I'm happy to take questions.
46:37OK. That was wonderful, Jackie.
46:40Thank you for that overview and I think
46:44please folks, feel free to put in.
46:48Uh, some comments or questions
46:50in into the chat and and while
46:52we wait for folks to do that,
46:55I you know, I think.
46:57It is having been in,
46:59in this domain at least on the ischemic
47:00stroke side for a number of years.
47:02It really is quite amazing to me
47:05the developments that the field of
47:08neurology has had just a really
47:10in the last five to 10 years
47:13with endovascular treatments and
47:15medications and thrombolytic therapies.
47:18It is, it is just incredible
47:21expanding the hours of of therapy
47:24that many of these patients can get.
47:28Um treatments,
47:29but they're still is I think a large
47:34group of patients who have I would
47:36say probably particularly Jackie on
47:39on on the ICH side that have really
47:42profound functional and and cognitive
47:45impairment after their stroke.
47:47And so I think we're where this work
47:50is particularly promising and the
47:52work that you're doing in ICH is,
47:55is the idea that something that can be
47:58applied really to the majority of people who.
48:00Who have had these strokes given the
48:03prevalence estimates you're finding
48:04that can be continued, you know,
48:06started in the acute setting,
48:07potentially continued to over
48:09rehab and has the potential to to
48:13perhaps improve functional recovery
48:15is is really I think a sort of
48:19a novel line of investigation.
48:22The.
48:22You know, a couple of specific questions.
48:25One is I think another another
48:29sort of rationale for embarking
48:31on this investigation given what
48:34we've observed on the ischemic
48:36stroke side is that there's a lot
48:39of overlap mechanistically between
48:41aschematic stroke and ICH in terms
48:44of atherosclerosis mechanisms.
48:46And I think that played a role
48:48into the analytic strategy that
48:50you engaged in on that original
48:52cross section looking at.
48:54Sort of a layered analysis with hypertension.
48:56I wonder if you might make might
48:58speak a little bit more about,
48:59about that and the challenges of
49:01looking at that,
49:01some of those founding factors.
49:03Yeah. And yeah, no, it's definitely.
49:06I I agree with all that you said and I
49:08think that you're right that it really
49:11all comes back to sort of small vessel
49:13ischemic disease which is not intuitive
49:15and and and certainly not something
49:19that I would have thought was at all.
49:22You know I think when when internist think
49:25about hemorrhagic stroke or particularly
49:27you know interest cerebral hemorrhage
49:29as a stroke subtype we think about
49:31aneurysm rupture you know brain tumor you
49:34know I think it it takes a little bit.
49:37On. More education, I think,
49:41than we get in our basic internal medicine
49:44training to really understand the nuances,
49:47but it's it's really fascinating
49:49to see all you know it all of the.
49:53The similarities in pathophysiology between
49:56a schematic stroke and hemorrhagic stroke,
50:00or particularly IC H and so I think.
50:04To your point,
50:05it really does sort of lend itself.
50:07It's sort of shocking when you think
50:10about it that that this hasn't been
50:13done and that actually so if you,
50:15you know for a lot of studies that have,
50:16you know,
50:17brain hemorrhage was excluded
50:19and then I think that.
50:22As far as the sort of process of teasing
50:25it out since hypertension is such.
50:31Sort of a main risk factor for IC H and we
50:35spend a lot of time trying to figure out.
50:39You know how,
50:40how it's impact might not be
50:43over represented,
50:44but also not wanting to completely
50:46eliminate it because we know that it has
50:48a complex relationship with sleep apnea.
50:50And so as you as you mentioned
50:53we did several analysis.
50:55It's also worth noting that the
50:58Berlin questionnaire includes
50:59hypertension as one of the main
51:01components of the questionnaire itself,
51:03and so on.
51:06What we ended,
51:06we did one analysis actually
51:08where we just removed hypertension
51:09from the Berlin questionnaires.
51:11We rescored them and then ran the
51:14analysis because we we wanted to sort of,
51:16you know,
51:17play a worst case scenario for hypertension,
51:19best case scenario for the connection and
51:22it's still remained a significant finding.
51:25And so I think that the the challenge
51:27for me and I think for a lot of us is
51:29that there is no standard approach
51:30to this and you want to make sure
51:33that what you're finding makes
51:34physiologic sense and is plausible.
51:36Um,
51:37but but there's no manual for how to do it.
51:40Or or you know,
51:41which which analysis is going
51:42to is going to say, OK, see,
51:44this is definitely accurate,
51:45but I think we can safely say that
51:47we're really on to something here,
51:48even when we exclude hypertension.
51:51OK, so there's a several
51:54questions in the chat.
51:56The first is from Doctor Hoffman,
51:59which is an interesting question.
52:01So any concerns about increased
52:02intercranial pressure with CPAP
52:04in an acute stroke setting?
52:07So I I've had the same,
52:09I've had the same thought.
52:10I I think the answer is that it's
52:14probably not high enough pressure
52:16and we're and we're not dealing.
52:19I I don't actually know
52:20exactly how to answer that.
52:21I think it's an excellent question
52:22and I think that the general,
52:24the consensus is that no.
52:28That the pressures are
52:29probably not high enough,
52:30but I can't say that this has
52:33actually been studied and to to my
52:35knowledge I'm not sure if if, Kevin,
52:38I see you're you're on the call or if
52:40anyone else has an answer to this.
52:43And I mean, it's a,
52:44it's a good question,
52:45but it turns out actually ICP is
52:48virtually never important Q stroke.
52:51And it just it just in terms of
52:53prevalence it's only in the hyper
52:55acute setting and the most sort
52:57of you know in the most severe or
52:59largest infarcts or hemorrhages that
53:00ICP ends up being an issue and and
53:02those patients either just going to
53:04be very small population or going to
53:06be excluded from all these studies.
53:09It would be an exception
53:10in all the other patients that just,
53:11it turns out not to matter
53:13at all. OK, thanks. That makes sense.
53:15So basically if it if it is a concern,
53:17the patients probably intubated anyway
53:19and by the time if they are able to be
53:23extubated and potentially use CPAP,
53:25it's it's probably not an issue anymore?
53:30There's another question, Jackie.
53:31And so based on your observations,
53:34when would you consider an ideal
53:36time to start PAP therapy in,
53:38in post stroke patients?
53:40So I think that's sort of
53:41the $1,000,000 question.
53:42And I think probably 1 component,
53:45which I didn't speak much about today
53:47but is vitally important, is adherence.
53:50I think you want to find the best
53:53balance on like when is it most tolerant,
53:55like when will the patient tolerate it?
53:58Because you, you know,
53:59even if there were some
54:00marginal benefit to starting it,
54:01like the day they come in for their stroke,
54:04say, you know, if they're totally
54:07overwhelmed by everything that's
54:08going on or they have some new.
54:10Um, functional limitation that they're
54:12aware of and they're, you know,
54:14there's a lot of sort of like.
54:16You know emotional, physical,
54:18cognitive components here.
54:20And so I think that like my overall
54:22answer is that the best time is when
54:25they'll use it if we think it's helpful,
54:28which I think the majority of us do.
54:30And that being said,
54:31I think that's a huge area of ongoing study.
54:35I mean particularly as far as its impact
54:37on outcomes what you know what timing
54:40is ideal for for outcomes and and I
54:42will put a plug in that we're we're
54:45also for rise up one of the things that I.
54:47I'm excited to be a part of is
54:49working on the qualitative component,
54:51kind of talking to people about
54:54their experiences in the study,
54:56with the path,
54:57with their stroke and kind of
55:00hearing a little bit more about.
55:03What?
55:04Like,
55:04what plays into that because
55:05they think adherence is key?
55:09I think those are all great points.
55:10I would just add you know we're
55:12a little bit further ahead on the
55:15ischemic stroke side than the ICH side
55:16at looking at the this timing issue.
55:18And I think some of the preliminary
55:22studies that have been done have
55:25suggested that the earlier the initiation
55:28of PAP therapy may be of theoretical
55:31benefit and the acute stroke setting
55:34to a brain and and and an ischemic
55:36penumbra of the brain that may be
55:39particularly susceptible to some of the.
55:41Physiologic sequelae of sleep apnea
55:44and that was that's sort of the.
55:46But at the same time to Jackie's point you,
55:48you have you have patients who
55:50are acutely stroked and might have
55:52difficulty adhering to or tolerating so.
55:55So I think that that time point is
55:57needs to needs to be worked out
55:59but the thing about you know having
56:01this as a complementary acute stroke
56:04therapy is I think very appealing.
56:07Umm.
56:10Let's see here.
56:12So Kevin, do you want to
56:14weigh in with your question?
56:15I'm just curious, Jackie,
56:16there's great, great talk.
56:18I may have asked you this before
56:20in the past, but good to just,
56:22let's just forecast the future of touch,
56:24let's say rise up and sleep,
56:25smarter neutral studies.
56:26How are we going to think about
56:29interpreting those findings
56:30and what's next for,
56:31what does that mean for sleep
56:33and also PAP and and stroke?
56:36Yeah, I know. I mean, I think that's
56:38like we have to kind of prepare that.
56:40That's probably going to happen, right?
56:41And then what do we do?
56:43What does it mean?
56:44It's sort of like all the sepsis
56:47and steroid trials.
56:48I think I'm going to say that it
56:50will probably then come down to the whole,
56:52I didn't talk too much about
56:54the phenotyping today.
56:55I think maybe for another talk I would
56:57love to to do that when I have some data.
56:59But I think that it may come down to an
57:04issue of sort of like of means versus sub
57:08you know subgroups and that there there
57:12may be certain sleep apnea phenotypes.
57:17Or or maybe sort of cardiometabolic
57:19phenotypes that do respond and and others
57:23who don't who kind of dilute the pool.
57:26And so, you know,
57:26it's always, it's tricky.
57:29It's it's tricky when it happens,
57:31but I I don't think the game
57:32is over and I do.
57:33I think we all kind of
57:34firmly believe that it.
57:35You know if sleep apnea
57:37causes is a cause for stroke,
57:39how can treating the sleep apnea with the
57:42world's most effective treatment not?
57:44You know,
57:45actually help them in some way or at least
57:47prevent recurrent strokes or cognitive,
57:49you know,
57:51cause, you know,
57:52improve cognition and things like that.
57:56Very good. Thank you very much
57:57for excellent presentation and
57:59and the great questions everyone.
58:00I think we're going to wrap up now.
58:02There's another conference
58:04immediately following this.
58:05Have a great rest of your week,
58:06everyone. Thanks everyone.