Sleep SMART Trial - Devin Brown and Ronald Chervin

October 26, 2021

ID7074

To CiteDCA Citation Guide

00:00To introduce sleep smart,
00:02Smart is a randomized clinical
00:04trial assessing whether treatment
00:06of obstructive sleep apnea shortly
00:09after an acute ischemic stroke or
00:12high risk TE reduces the risk of
00:14cardiovascular events and improves
00:16functional outcomes representing sleep.
00:18Smart today are Doctor Devin
00:20Brown and Doctor Sherman.
00:22Dr Brown is a professor of
00:24neurology at University of Michigan
00:26Medical School Doctor Sherman,
00:28who is also a professor for
00:29Alaji and a director.
00:30Sleep Medicine at University
00:32of Michigan Medical school.
00:34Thank you very much for joining us today.
00:37Thank you so much for the invitation
00:39and it's really wonderful to be
00:41able to stop by for a visit and talk
00:44with you briefly about sleep smart.
00:46As we already said,
00:47Ron Sherman is also on the line,
00:50and so I'm sure he's happy to answer
00:52any of your most difficult questions
00:55will save those for him for at the end,
00:58Sharon's presentation was was very
01:00heartwarming at and her house is
01:02also quite lovely compared to mine,
01:04so apologies for the background.
01:07And perhaps for the less
01:11heartwarming presentation.
01:13And so let's start just by talking a
01:15little bit about why in the context
01:18of caring for a stroke patient,
01:20you would want to even care
01:21about obstructive sleep apnea.
01:22There are so many other things to
01:24consider to worry about to focus on.
01:26Why are we obsessed with
01:29obstructive sleep apnea,
01:30well, obstructive sleep apnea
01:32is very common post stroke.
01:34It is approaching the prevalence of
01:36hypertension, so it's up in the 70s.
01:39So when you see a stroke patient,
01:41the chances of that person.
01:42Obstructive sleep apnea are extremely common.
01:46We know that obstructive sleep apnea
01:48is an independent risk factor for both
01:51incident stroke and recurrent stroke,
01:53and it's also an independent risk factor
01:55for poor outcomes following ischemic stroke,
01:58including both functional
02:00and cognitive outcomes.
02:03So what are the links there?
02:05How does obstructive sleep
02:06apnea potentially cause stroke?
02:08How does it potentially cause
02:10poor outcomes after stroke?
02:12Well, there are lots of different.
02:13Possibilities first.
02:15Sleep apnea causes elaboration of
02:18free radicals of khyle 6 E selected.
02:23These things can promote Atheros sclerosis.
02:25It promotes deleterious cerebral
02:27hemodynamics and through
02:28multiple different mechanisms,
02:30including platelet activation
02:32and increased EPO,
02:34and decreased fibrinogen increases
02:36hypercoagulability and any of these
02:38three factors alone or in combination,
02:41can increase the risk of
02:42both incident and recurrent.
02:44Stroke and then following a
02:45little bit of a different pathway.
02:48Sleep apnea again through all
02:50of the different physiologic
02:51changes that it can cause.
02:53May end up producing angiogenesis,
02:55dendritic and axonal
02:57sprouting and synaptogenesis,
02:59and these factors can result
03:02in poorer stroke recovery.
03:05So,
03:05given that we have both of these two
03:07very important potential outcomes,
03:10recurrent stroke and stroke recovery,
03:12which should we target in a
03:15trial looking at treatment of
03:17obstructive sleep apnea poststroke?
03:19Well,
03:20our approach was really that we wanted
03:22to have our cake and eat it too,
03:24and so this says you can't
03:25have your cake and eat it too.
03:27That's obviously being stated by
03:29somebody who doesn't understand what
03:30you're supposed to do with cake,
03:32so we took this approach that we wanted to.
03:36Test both of our hypotheses that CPAP
03:38could improve prevention and it could
03:41improve recovery and within sleep.
03:43Smart participants are enrolled as if it
03:45is a single trial they've taken through.
03:47The protocol is if it's a single trial,
03:49but then at the time of analysis,
03:51which hopefully will come at
03:53some point in in several years.
03:55It then breaks down into
03:57really two separate trials,
03:58one where all participants are contributing
04:01their data to the prevention outcome,
04:03and those include both the high risk Tia.
04:06Of which there are very few,
04:08and the ischemic stroke patients,
04:10and they have to be enrolled
04:12within 14 days of symptom onset.
04:14But then to answer the recovery aim,
04:16we use only a subset of
04:18the enrolled participants.
04:19Those who had an ischemic stroke
04:21within seven days of consent,
04:23and those who also have to have had
04:26an NIH stroke scale of at least
04:27one at the time of enrollment,
04:30because, otherwise,
04:30how are you going to be able
04:32to note that there has been an
04:34improvement in in their recovery?
04:37So the design of sleep smart is that
04:39it is a late phase multicenter trial.
04:42The control group is usual care,
04:45so it's usual care.
04:47Plus automatically adjusting
04:48CPAP versus usual care alone.
04:51We could have designed this to have an
04:53active control will not active control,
04:55but a placebo control using sham
04:57CPAP and it's something with
04:59which we have experienced.
05:00But it really would have complicated
05:03our design substantially and
05:05it would have advocated the.
05:07Possibility of are using a run at night,
05:09which is a key part of our protocol design,
05:12and so in the face of knowing that
05:14we are using open label treatment,
05:16we used a probe design where the
05:19outcome assessors are masked to
05:22randomization assignment and
05:23then again as I said before,
05:24this is really a secondary prevention
05:27trial with an embedded recovery trial.
05:32This shows how a participant
05:34goes through the protocol.
05:35So after consent and
05:37baseline data collection,
05:38the first night is allocated to
05:41sleep apnea testing with an ox T3
05:44sleep apnea device and then to have
05:46qualifying obstructive sleep apnea.
05:48The Respiratory Event index has to
05:50be at least 10 and half of those
05:52events cannot be no greater than
05:55half of them can be central events,
05:57and then the person moves on to
05:59the second night where he or she.
06:00Essentially gets a taste of C.
06:02Pap gets to try it out in the run and night,
06:05and if that subject uses C PAP
06:07for release cumulatively 4 hours
06:09during that night and also does
06:11not exceed 10 for the central
06:13apnea index read off of the device.
06:16So meaning therefore the person did
06:18not have treatment induced central
06:20sleep apnea and the participant is
06:23willing after that one night of
06:25exposure to see PAP to have a 5050
06:27chance of intervention versus control group,
06:30then that person.
06:31Is eligible for randomization and
06:33receives again either automatically
06:35adjusting CPAP plus best medical
06:37therapy versus just best medical
06:40therapy alone and then we follow
06:42the subjects for three months for
06:44the recovery outcomes and then six
06:46months for the prevention outcomes.
06:51We were asked to cover a couple of
06:53different topics during this brief talk,
06:55and so I'm going to move on to enrollment
06:59criteria and how we we conceptualize those.
07:01And we're going to try to highlight some
07:03of the questions that were asked of us.
07:06So the enrollment criteria really
07:08quite broad in sleep smart.
07:09We're trying to have a generalizable trial.
07:11We're trying to have a treatment
07:13that potentially can help the
07:15most number of participants.
07:17So there the inclusion criteria are really.
07:19Very broad.
07:20If you've had an ischemic stroke or high
07:22risk TA in the prior 14 days in year,
07:24at least 18, and you're asleep smart site,
07:27you're essentially eligible from the
07:30inclusion side of things we have.
07:33For what I'm going to describe is 4
07:35categories of exclusion criteria.
07:37The first are really the general ones,
07:39so if you have somebody who's a
07:42pregnant woman incarcerated and
07:43can't sign our own consent,
07:44that somebody who you're
07:46going to want to exclude,
07:47and if it's somebody who could
07:49not perform all of his or her.
07:50Activities of daily living
07:52prior to the stroke.
07:54Then that's also someone
07:55who would be excluded.
07:57The next category are the
07:59CPAP specific related issues.
08:01So if you are on currently on mechanical
08:03ventilation or if you have a tracheostomy,
08:06you're not going to benefit from
08:07C PAP and so you're excluded.
08:09And then if you've used C PAP
08:11in the last month,
08:13we have a concern that if your
08:14randomized so the control group
08:15you're going to go home and use your
08:17see PAP and therefore crossover,
08:18so you're excluded for that as well.
08:21The third category are things that we
08:24think potentially could make CPAP riskier.
08:26It's very low.
08:28Risk treatment,
08:29but there there are some factors that
08:31may increase risks and so those include
08:34things such as bullous lung disease,
08:37pneumothorax having hypo tension
08:39that's so significant that you're
08:41requiring pressers at that time.
08:43If you've had massive epistaxis.
08:46If you have a possible CSF
08:48leak or Numa cephalus,
08:49or if you've had any kind of bone
08:51off procedure where the bone has
08:53not been replaced on the head,
08:55then C Pap maybe a little bit more risky.
08:58In those participants,
08:59and therefore they are excluded.
09:00We also have a category for the site P.
09:02I feeling like there's some other
09:04entity that increases the risk
09:06of C PAP and so we allow for of
09:08course the judgment of the local
09:10teams to decide this is not a good
09:13idea for our for our patient.
09:15And then the 4th category really is
09:17something that makes it really unfeasible.
09:20So for instance,
09:20if the participant or if the sort of the
09:23patient is using oxygen supplementations
09:25greater than four liters per minute,
09:27you can't believe.
09:28That into our CPAP machines,
09:29and therefore it's really unfeasible
09:31and then if that person is on some type
09:34of precautions, contact precaution,
09:36respiratory precautions,
09:37we don't want to cross contaminate
09:40with our equipment and infect
09:42another participant so it really
09:44becomes unfeasible.
09:45Switching gears a little bit,
09:47we were asked to talk a little
09:49bit about the stroke physicians
09:51versus the sleep positions and
09:53how those interactions occur.
09:55We've had some comments from potential
09:59sites where they have said is CPAP
10:01really safe for stroke patients?
10:04So I'm concerned that if my person
10:05if my patient is enrolled and then
10:07randomize the intervention group,
10:09that CPAP could potentially cause harm
10:11that is most commonly said by a sleep.
10:14I started a stroke.
10:15Position if it is,
10:16if it said and on the flip side,
10:18there are some sites where they'll
10:20come back to us and say how
10:21can you withhold CPAP after you
10:22know that the patient has been
10:24diagnosed with obstructive sleep
10:25apnea by randomizing that person
10:27to the control group that is more
10:30commonly said by a sleep physician.
10:33And so overall,
10:34we really feel that we are in a
10:36position of clinical equipoised with
10:37respect to C PAP for stroke patients.
10:40We don't know whether CPAP will
10:43help harm or essentially do
10:45neither for our stroke patients.
10:47There have been no definitive
10:49randomized controlled trials for stroke
10:51patients for stroke outcomes that have
10:54shown anything is improved by CPAP.
10:56So we feel comfortable with
10:57holding it from the control group,
10:59and there there's precedence for this.
11:01There have been numerous.
11:03Randomized controlled trials that
11:04have enrolled either patients
11:06with cardiovascular disease,
11:08such as Save Rick Ads or SIRKAS,
11:11or that have enrolled lots of
11:14participants with severe sleep apnea,
11:16such as apples where patients are
11:18randomized to a control group or,
11:20in the case of apples, to a sham control.
11:25So other investigative teams,
11:28other funding agencies,
11:30other peer review panels have found
11:32this to be completely ethical
11:34and not have any concern.
11:37There's also the 2017 U.
11:39S Preventive Taskforce report that
11:41helped inform our our decision making
11:45at the time that we were designing
11:48sleep smart and and proposing it
11:50for the first time that states that
11:52there is no established benefit
11:53of C PAP for any health outcome.
11:55This is just.
11:55In the general population,
11:56not even specific to stroke aside
11:59from the modest improvement in
12:01sleep related quality of life,
12:02and the more recent U.
12:03S preventive taskforce doesn't say
12:06anything that would compel us not to
12:09randomize participants to a control group.
12:11We were also asked to talk a little
12:13bit about crossover so crossovers
12:15where you have a control person who
12:17someone who's randomized to the control
12:19group who then wants to use CPAP.
12:21So when that does occur and it's
12:23not something that we thought
12:25would be very common.
12:26Based on our preliminary work
12:28and based on prior CPAP trials,
12:30pilot trials among stroke
12:32patients if that does occur,
12:34then the clinical team should absolutely
12:37feel free to refer the participant
12:39for sleep apnea testing for sleep.
12:41Get me a treatment in the clinical realm,
12:43it usually takes some time for that
12:45to be available to the participant,
12:47so it may actually.
12:50Push the see PAP treatment for clinical
12:52care outside of the even six month window.
12:54By the time the person is able
12:56to get tested and treated and
12:58have a C Pap in his or her home,
13:00but the research team, we would suggest
13:03not help facilitate that process.
13:05It is a protocol violation for a
13:07control participant to start using
13:08C PAP so it has to be reported
13:10as such and in the analysis,
13:12at least in the intent to treat
13:14component which is our primary analysis.
13:16The control participant who starts
13:18using C PAP will be analyzed.
13:20As a control participant.
13:23Crossovers from control.
13:24Two intervention or to to CPAP use
13:27have been very uncommon in in sleep
13:30smart so far it's been around 2%.
13:35So what about anticipated challenges?
13:38Well, we knew that recruitment
13:39would be an issue.
13:40Recruitment is an issue for every
13:43randomized controlled trial.
13:44CPAP adherence is an issue
13:46for every CPAP related trial,
13:49but some of the things that we did
13:51not anticipate having difficulty
13:52with included a global pandemic.
13:55We did not presage that,
13:57and having difficulty achieving
13:59in window outcome assessments
14:01has been much more challenging.
14:04Then we had anticipated we have
14:06more missing data at the three
14:08month TIMEPOINT for the modified
14:10Rankin which is our primary for
14:12that aim than we had anticipated.
14:14We did try in the design of sleep smart
14:17to prepare for some of these challenges.
14:20So for instance we built in telephone
14:23outcome assessments from the onset
14:25that was always allowable and sleep
14:27smart even pre COVID and most of our
14:30outcomes can be assessed by telephone.
14:32There are only a few secondary outcomes.
14:34Exploratory outcomes that cannot,
14:36but most of them can.
14:37We really tried to be very careful and
14:40intentional about the selection of our
14:43outcome assessments to make them as
14:45short as possible and when possible,
14:47to allow something to be conducted by phone.
14:51We also created a lot of tools for
14:54site teams to be able to reach out to
14:56participants in case there were any
14:58issues trying to achieve outcome assessments.
15:01So we built in a place in the
15:03back of the consent form.
15:04For instance,
15:05where lots of contact information,
15:07alternative contact information
15:08for the subject,
15:10alternative contact information for partners,
15:12friends,
15:13family members could be
15:14documented and then referred to.
15:16We created several letter templates
15:18for sites to use to reach out to
15:21subjects about scheduled appointments.
15:23Missed appointments unable to reach
15:25those types of things we've developed.
15:28A slide set that sites can
15:31use to help educate teams.
15:34Clinical teams,
15:35including nurses about sleep
15:37smart and we created a document
15:41that provides our sort of answers
15:44to potential difficult patient
15:46questions at the time of enrollment,
15:49and we also of course,
15:50built in Tele Med telemedicine
15:52approach to outpatient.
15:53Management of CPAP,
15:55which in COVID has been very advantageous.
16:01But despite the challenges,
16:02there remain lots of hope.
16:04There's hope because the of the vaccine,
16:07which hopefully will assist teams and
16:11getting back to their usual state when
16:13it comes to coordinator coverage.
16:15Respiratory therapy support,
16:16but mostly our hope comes from
16:19our sites and the sites have been
16:21doing a fantastic job despite the
16:23pandemic in the face of a pandemic,
16:26we are really grateful to every site.
16:28There are some sites,
16:30as you see who are randomized.
16:32In the 50s and the 60s,
16:34number of participants,
16:35which is fantastic.
16:37I would like to give a little
16:39shout out to two of your sites.
16:40North Shore with 16 and Yale
16:44with nine Randomizations.
16:46We're very grateful to you
16:47for all of your work.
16:49You also have Hartford and Staten Island,
16:52and we are grateful for those sites as well.
16:57And then just looking at by RCC
16:59and you see that some are CC's
17:02are just going gangbusters.
17:04Some are not participating in sleep
17:06smart that is very few of them.
17:08And then I've outlined Yale doing very
17:12well here somewhere in the middle.
17:16And so I thank you very much for again,
17:18the invitation and for your attention.
17:22And again, Ron is is available to answer
17:24any difficult question that that you have.
17:30Thank you very much for that Devin.
17:32Uhm, I just had a question
17:35about a trial powering.
17:37Whether it was parked for both
17:39the cardiovascular events as
17:40well as the recovery outcomes.
17:43Yeah, no, that's a good question.
17:45So we did look at power calculations
17:48for both and we anticipate that a
17:51certain percentage of the total will
17:54be available for the recovery outcome.
17:57And it turns out that we are we have a
18:00higher proportion than we had anticipated,
18:02so we we think that those two are
18:04kind of going to ride along together
18:06and that by the end we should have
18:08a sufficient number in both groups.
18:11Fantastic in such an innovative
18:13innovative trial design. Thank you.
18:24Alright, I think those are the questions.
18:26Well, thank you so much for joy.