ARCADIA-CSI Trial - Maarten Lansberg and Ronald Lazar

October 27, 2021

ID7077

To CiteDCA Citation Guide

00:00To introduce you and we will
00:03be hearing about Arcadia,
00:05CSI and ancillary study of
00:07the parent trial Arcadia,
00:09which aims to determine the effect
00:11of anticoagulation on poststroke
00:13cognitive function and incidents of
00:15silent infarcts representing Arcadia,
00:17CSI or Doctor Martin Lansberg
00:19and Doctor Ronald Lazar.
00:21Dr Lansberg is professor of neurology
00:23and neurological sciences at
00:25Stanford University Medical Center.
00:27Dr Lazar is professor of
00:30neurology and neurobiology.
00:31At University of Alabama at
00:33Birmingham School of Medicine.
00:34Thank you.
00:36Thanks Richard for the introduction
00:38and good morning everyone,
00:40and I wanted to share a little bit
00:43of what we've learned in setting up
00:45an ancillary study today and then a
00:48little bit of background on on on
00:51the silent infarcts after stroke.
00:53And then Ron will be talking
00:55mostly about cognition so.
01:02Background is that in full 2017 we
01:07were finishing up to diffuse 3 trial.
01:10It had stopped early and.
01:14After the Dom trial had finished
01:17enrollment and we had to take an
01:21early look at our data and at that
01:24point we wanted to use the funds of
01:27diffused three to continue studying
01:29large for patients with the remaining
01:31funds that we had in the study,
01:33but that that was not an option because
01:36we had to put in a new proposal for
01:40the diffuse for large course study,
01:42we couldn't use the unused funds and then.
01:45Talking to Scott Jennings at that
01:47time got some good advice and
01:49sort of like off the cuff said,
01:51well, you know,
01:53sometimes these ancillary proposals and have
01:55a better chance of getting funded and so.
02:00I, I think around the same time,
02:02or maybe at the same meeting Kevin
02:06and I were discussing various trial
02:09ideas and one thing that came up
02:12was what about the Arcadia trial?
02:15And they're not doing any
02:17imaging in the Arcadia trial.
02:19Perhaps we should think about an
02:21ancillary study to the Arcadia trial,
02:23and then, if we're doing imaging.
02:25And what about Viet being
02:27interested in cognition,
02:28but about also looking at cognition and?
02:32And that's how the idea was born, uhm?
02:37And we talked to the Arcadia trial P
02:39eyes to see if they would be supportive
02:41of this ideas and and they were.
02:44And then we were lucky to assemble a team
02:47of really outstanding experts in cognition.
02:51Wrong loser will be talking in
02:53in a few minutes. Imaging marks.
02:55Winter Market Center radiologist
02:58and statistics.
02:59George Howard,
03:00who is now spent his whole
03:02life looking at them.
03:04I have to give Myologie in statistics
03:06in stroke and come together with their
03:10Arcadia trial pies we got on the phone
03:14weekly to get a proposal written.
03:17And to our vision was to get
03:19large scale data of the effect
03:22of anticoagulation on two things,
03:24silent infarction, cognitive decline.
03:25And we thought the results will
03:28be important regardless of the
03:30results of the parent trial.
03:31That would provide deeper insight in
03:33the incidence of silent infarcts,
03:35the slope of post stroke,
03:36cognitive decline,
03:37and the relationship between these two,
03:40and that it would set the stage for
03:43future trials to prevent both silent
03:45infarction and post stroke cognitive decline.
03:48As Ron will tell you,
03:49we felt that you know these are
03:52aspects of stroke that are understudied
03:55and but very relevant to patients.
03:57And so this vision and rationale
04:00was really how we try to sell the
04:03ancillary study in our brand proposal.
04:07And then we come.
04:11Little bit of background about
04:13RPDR and this apparent trial.
04:15It's apixaban versus aspirin for
04:17secondary stroke prevention.
04:18You'll hear more about that later
04:20today from the from the trial piece,
04:23and they plan to enroll 1100 patients
04:25with cryptogenic stroke and atrial
04:28cardiopathy plan of 2 1/2 years of
04:30enrollment and a median of three
04:33years of follow up.
04:34There was no data on cognitive
04:36and or imaging outcomes,
04:37and So what we would add was these
04:39key cognitive in or imaging outcomes.
04:41We wanted to integrate seamlessly
04:43with Arcadia study visits.
04:45So there would be easy for the
04:47coordinators and the patients and low
04:49barrier for entry and and low burden
04:52on the patients and investigators
04:54and hopefully high impact if we
04:57could show any benefit either on
05:00silent infarcts or on cognition.
05:04So did bit of background on silent inverse.
05:06Very common prevalence 30 to
05:0950% in studies instance up to
05:1390% every year after TA.
05:16And there's data that they
05:18accelerate cognitive decline,
05:19which shown here in the graph
05:21on the left for patients with
05:23silent infarcts that follow up.
05:25Hey, decline in their cognitive function,
05:28and whereas those without
05:30a sign in first do not.
05:34So next I wanted to go on to sort of,
05:37UM, the lessons learned better than
05:39to talk about our protocol too much.
05:42I thought that for those of
05:44you who may be interested in.
05:47Designing your own ancillary study
05:49that it might be helpful to talk
05:51about what we learned in the process.
05:53And so the first thing is that it takes time.
05:57In October 17, after Kevin
06:00and I talked about the plan,
06:03we wrote the first draft of
06:05the concept synopsis.
06:06And two months later, be it in.
06:12Submitted this concept proposal to
06:14the prevention working group for
06:16our first meeting with them and
06:18received feedback on the proposal,
06:20and we're lucky that after a single review,
06:23we were allowed to submit it
06:26to an indie S for review.
06:28So we put a budget together and then.
06:34Updated the concept synopsis a
06:37little bit and submitted it to
06:40NIS for ESC review in March.
06:43In in we got also a positive,
06:47so again lucky that we got a
06:50positive response from NMDS.
06:51They said yes, you're allowed to
06:54submit your brand 2 NDS, and so we did.
06:57We then went to work and got the
07:01grant written and submitted a couple
07:042-3 months later, in June 1st, 2018.
07:09Half a year later, December 18,
07:12we received a summary statement from an end,
07:14yes, which looks favorable.
07:15But of course you don't know
07:17about funding at that point.
07:19And February 19,
07:21we got funding approved by Council and.
07:25We're excited that we could get started
07:29thinking about implementing the trial.
07:31We wrote our protocol and submitted it
07:36for the central IRB in March of 19.
07:39And then went to start working on
07:42contracts and grants with all the
07:44sites and be at the first page in the
07:47study about half a year later in November 19.
07:50So for those of you who
07:53are thinking about this,
07:55I think the big picture takeaway
07:57is that from first draft of the
08:00concept synopsis to first patient in
08:03study is two years and that is with
08:05everything working flawless Lee and
08:09getting funded on the first round.
08:13So it in be in for the for the long hold.
08:16This is not a it's not even though
08:18it's an ancillary study.
08:19It's not something that comes.
08:22Easy or quickly.
08:25A second learning point is that it
08:27isn't cheap. Total direct costs are in.
08:31It's always a little bit difficult
08:33with the stroke net studies to to
08:35save what's direct and indirect
08:36with somewhere in the five to $6
08:38million range for for this study,
08:41and I wasn't really aware of that going in,
08:44I thought we're adding imaging and
08:47we're adding cognitive testing,
08:49but we can do that relatively inexpensively,
08:52and. It it adds up. Why does it add up?
08:56Well for the ancillary study,
09:00the way they set up,
09:02you need separate contracts with each site.
09:05You need to separate curb.
09:07There's a separate database at Mercy
09:11that is maintained for this study.
09:14And so and then of course there's
09:17a separate trial management group
09:19of the pies and study coordinators
09:22to run the trial so.
09:25Altogether, and the amount adds up.
09:29And then there there's reimbursements
09:30to the site.
09:34So that is the the second learning point.
09:36The third one is that even though we
09:38kept eligible eligibility criteria
09:39as simple as possible, the inclusion,
09:41pretty aware that patient needs
09:43to be randomized in Arcadia.
09:44Need to be able to undergo MRI able
09:47to provide consent in English and
09:49score 0 to one on the NH language
09:52at the time of Arcadia enrollment,
09:54so they could undergo cognitive testing.
09:58So I'm very simple.
10:00Exclusion criteria really shouldn't
10:02exclude too many patients.
10:04Get the protocol simple,
10:06so simple MRI protocol,
10:0850 minutes scanning time 30 minutes
10:10total time can use them right from
10:13the index stroke at baseline as the
10:16baseline imaging study for enrollment.
10:18So no really,
10:19for most patients just one MRI
10:21stand scan at the end of the day,
10:24Arcadia participation.
10:25And so despite this,
10:28enrollment hasn't been easy and we
10:30see here that and our involvement
10:32is up to 203 which were.
10:35Super excited about it and then
10:37thanks to all the sites that are
10:40here on the call for enrolling.
10:42But the parent trial right now has
10:45enrolled over 700 patients and we
10:47were hoping to enroll about 50%.
10:49We thought it would be easy during
10:51roll 50% of of total enrollment
10:53in the parent trial.
10:54So we we had hoped to be
10:56somewhere around 375 now.
10:58And why are the Orange
10:59line is our involvement?
11:01The grey line you can ignore for now,
11:04so why is it difficult to
11:06enroll in ancillary product?
11:07Well.
11:09One reason is that the parent trial
11:12added additional sites and and some
11:14of those sites are in Canada and it
11:16was challenging for us to go to Canada.
11:19That's a small reason,
11:20but that's one of the reasons.
11:22Two is that we hadn't fully
11:26anticipated patients.
11:27The number of patients being off
11:29study truck in the parent trial,
11:30and those are no longer eligible
11:33for involvement in CSI because we
11:35cannot study our primary outcome.
11:37Patients need to consent separately
11:39for this ancillary study and
11:41and for the parent trial,
11:42and that is some additional burden for
11:44the patient and some recent patients
11:46may not participate and some sites
11:49declined to participate because of
11:52financial constraints that they're
11:53under and making it hard for them
11:56to take on both the ancillary study
11:58as well as their parents study.
12:00Even though we did our very best
12:03to maximize reimbursement to
12:05sites and and and minimize the.
12:09Overhead.
12:11And then four lesson is that it's
12:13hard to fully integrate with the
12:15parent row if the ancillary crowd
12:17goes in as a separate proposal
12:20by different investigators.
12:21So even though we've worked very
12:23closely with the Arcadia Pies,
12:25and they've been super supportive,
12:27we remain two separate study teams,
12:29and we have slightly different goals,
12:31and rightfully so.
12:32The parent trial wants to finish
12:34their trial and wants to finish their
12:37their enrollment as soon as possible,
12:39and so they need to do everything.
12:41Like going to Canada,
12:43add additional sites to to make
12:47sure that their trial is successful.
12:50And we are somewhat.
12:54We we we are very mindful
12:56that by our ancillary trial,
12:58we want we don't want to impede
13:00their enrollment in any way
13:02by overwhelming a patient,
13:04who then may say, well, you know,
13:06this is just too much.
13:06I'm not gonna even enroll in the
13:08Arcadia parent trial anymore.
13:12And uhm, we, but what I didn't realize.
13:15Also, I think going into this is that,
13:17uhm, we, we would have separate
13:20contracts with each side,
13:21so the side would have a contract for Arcadia
13:24and they would have a contract for Arcadia.
13:26CSI because of that we have
13:29slightly different sites, right?
13:30Some sites may decide to do Arcadia,
13:32but not CSI.
13:33And because we have different sites to be,
13:36we end up having slightly different webinars,
13:39different newsletters. So it is.
13:42Uhm, if the ends later study is run by by
13:45a different investigative team, it is.
13:48We can work closely together,
13:50but we cannot fully integrate,
13:52which may make enrollment and easier
13:55and make it a more cohesive project.
13:58And so this is our whole team.
14:02We couldn't do it without
14:03and I'm going to stop,
14:04share and hand it over to wrong
14:06to talk more about the cognitive
14:07aspects of the study.
14:23So I I've been told there's a 5 minutes left,
14:26so cognition is usually the caboose on
14:28the train, and these presentations,
14:29but I've gotten used to it,
14:31so I want to thank the the organizers of the
14:35Spirit Group today to allow us to present.
14:39So I'm going to go through quickly about
14:41what we're going to discuss regarding.
14:43Your card and assessment scientifically
14:46and logistically, and I think the first
14:48thing we have to recognize for ourselves
14:50that cognition is not one thing.
14:52It's a rubric. It's memory, language,
14:55attention, executive function,
14:57perception, processing, speed,
14:58and so when we do a study,
15:00whether it's Arcadia, CSI, or any others,
15:02we have to ask ourselves if cognitions
15:05and outcome was it we're trying to study,
15:08and that if we're doing memory when
15:10not necessarily doing language,
15:11and so we have to be circumspect
15:13about what we want to.
15:14Pay attention to and.
15:15Secondly, cognition is highly
15:17sensitive to brain integrity,
15:19which means that you don't necessarily
15:21need to have a lesion on MRI in order
15:24to have an impact on cognition from
15:27an intervention or from disease.
15:29So we have to tell her the battery
15:31to the disease we're studying.
15:33If you're looking at focal stroke then
15:35you can study language and recognition,
15:38factual memory, practice, spatial perception.
15:42But if you're looking at multi
15:44focal or diffuse disease then you're
15:46dealing with more widely distributed
15:47functions such as attention,
15:49working memory,
15:50executive functioning reasoning
15:52and processing speed.
15:53So whatever you decide to do to choose as an
15:56outcome in terms of your cognitive domain,
15:59it should match the.
16:00Pathology that you're trying to study.
16:02So this is a slide from yells oh
16:05natural Alexandria Lanski,
16:07who chaired the Neurologic Academy Research
16:10Consortium on outcomes for neurologic.
16:15Effects from cardiac treatment
16:17and disease and what I want to
16:19point out to you from here,
16:21and I participate in this consortium
16:23is that we can use cognition either
16:26as part of an effectiveness trial,
16:28which would be an example of Crest age,
16:30where one want to see whether or not
16:33restoring perfusion would increase cognition,
16:34or it could serve as a safety outcome.
16:37You decide where does cognition belong,
16:39and sometimes it's going to
16:41correlate with the DWI image,
16:43and sometimes it's not going
16:44to correlate with a DWI image.
16:46So we have a wide range of possibilities
16:49about how cognition can exist
16:52in a cardiovascular trial.
16:54Who does the assessment in in these studies?
16:56Well,
16:57here are four examples that I've used
17:00over the years in the study above the
17:02amphetamine and hand cognitive recovery.
17:04After stroke we use nurses and both
17:07registered nurses and nurse practitioners,
17:10and accord C,
17:11which is the action to control
17:14cardiovascular risk in diabetes study.
17:16We actually used assistant clinical
17:18coordinators at about 50 sites around
17:21the United States that we trained.
17:23Then in the Sentinel study,
17:24which I'll talk about in a
17:26second which only had 19 sites,
17:28we use clinical notes are
17:30neuropsychologists and then for
17:31Crest two and Crest H and Arcadia,
17:33CSI. We're using centralized telephone
17:35based assessment because of the very
17:38large number of centers we want to do.
17:40If you want to standardize
17:42assessment in our judgment,
17:44it's best to do it from one central
17:46place so we can maintain the quality and
17:49and of the evaluations that were doing.
17:52So the battery that we're using for KDS I,
17:56which is the same as that fircrest
17:59two consists of being able to give as
18:01many words as possible to FA&S room.
18:03But this is a telephone based battery.
18:05How many animals can you name in a minute?
18:08Your ability to learn 10 words
18:10over three trials,
18:11a later recall of those 10 words,
18:13digit span and oral trail making.
18:16So the first question when the when
18:18the questions we wanted answered
18:19for ourselves is how sensitive
18:21is this battery battery in there?
18:23In the context of a telephone administration,
18:27so using the Crest two data,
18:30we compared the Crest data
18:33from these tests here.
18:36Letter fluency animal naming,
18:37word list learning and word list
18:40recall to the regards study at 30,000
18:43patient population based study of
18:45white and not for American blacks,
18:47largely in stroke belt and what we
18:49simply did is took the score from Crest,
18:51two subtracted from that demeans.
18:54Regards score in regards to any
18:56deviation and calculated Z scores for
18:58individuals from Crest two which was
19:00joke free and then match for age,
19:02sex,
19:03education and risk factor and
19:06what we found briefly.
19:07Is that if you take a look at the
19:09expected distribution which is regards
19:11and you compare the results across
19:14Crest 2 for Animal naming letter F only
19:16wordless learning and word list recall.
19:19You can see that there has been
19:22relatively little impact for
19:23naming for letters and and animals.
19:26Some impact on on on learning,
19:29but a significant decline in wordless delay.
19:32And these are data that coming out in stroke.
19:35So did this shows that there
19:37is some sensitivity to.
19:39To a telephone based battery from
19:41from at least carotid disease,
19:43which of course is not the same
19:46mechanism for Arcadia CSI,
19:47but the impetus in in looking at
19:50sound infarction that we could use.
19:53Comes from the Sentinel data,
19:55where we looked at how patients
19:58perform them cognitive testing before
20:00and after a Tavern and what we have
20:04here from 189 patients is this is
20:07the log transformed lesion number.
20:09This is disease score change before and
20:11after Taver 0 represents no change,
20:14a better score,
20:15a worse score and you could see
20:17the more the greater number of
20:18lesions the worst of cognition is
20:20and if you look at lesion volume.
20:22Essentially you get the same fining,
20:24so we know that in the setting of silent
20:28infarction in the absence of stroke,
20:30that cognition is a sensitive metric of.
20:34Of a silent infarction here.
20:38So in a in a brief one New York minute,
20:42that's kind of the way we're
20:44we're attacking cognition in the
20:46setting of sound infarction,
20:48and we'd be glad to answer any
20:50questions that you might have. And I'll.
20:53Thank you very much.
20:55Doctor Lutheran in doctor plants Berg,
20:57I believe there is a question
20:59in the chat by Shaggy.
21:03Great point about the consent
21:05process of ancillary studies.
21:07What is the barrier of having
21:09Arcadia CSI consent included
21:11in the Arcadia consent?
21:14And I just answered in the chat also,
21:17it's my understanding and perhaps there's
21:21people on the call from from Cincinnati
21:23who can answer this even better.
21:26That because the answer studies he
21:28knows are completely separate study,
21:29that is has its own review process
21:32in the CRB that separate consent is
21:35required and that it is logistically
21:38just difficult to integrate it.
21:40In in the Arcadia consent form,
21:42partially also because Arcadia patients,
21:44probably, you know,
21:45a lot of patients are consented in Arcadia,
21:48who will never get randomized,
21:49and we only look at the randomized patients.
21:53I don't know if Ben Plummer or anybody
21:56else has had more eloquent response. So,
22:01so Karen Fury asked a really
22:03important question about having
22:05English consent as a entry criterion
22:08for for getting into the study.
22:10And that raises a larger issue is
22:13because the test battery is also in
22:16English and and so I remember when we
22:20were going through the process of.
22:22The stroke net.
22:25Planning group that Ralph Sacco
22:27is exactly the same question,
22:29how come you don't have Spanish
22:31and it is definitely a limitation
22:33of our study and will limit the
22:36kind of conclusions we get.
22:38So try to translate the the battery.
22:41We're using that telephone base
22:43that came from regards and then
22:45into Crest too and and and and CSI.
22:47We've been trying to get funding
22:50to translate that battery and
22:51to validate it and and it's been
22:53a complicated process, but.
22:55When we have to do so,
22:56the limitation is acknowledged.
23:00Thank you another question from Walt Kernan.
23:03Is cognition a possible outcome
23:05for trials intended to gather
23:07preliminary data for novel therapies
23:09for secondary stroke prevention?
23:11If so, how many patients are needed
23:12to show typical treatment effect?
23:16Well, that time Walter,
23:17you always ask the hard questions and
23:20we don't have a a good answer to that.
23:24So our outcome in in Arcadia
23:26CSI is not that different.
23:29At the end of four years,
23:30but the slope and and so we we
23:34will only know at the end of the
23:36trial if the slope is separate
23:39far enough apart overtime whether
23:41or not we can get a treatment,
23:43you know that they did that.
23:44Martin present from Vermeer,
23:45was a population based study.
23:47And these patients, probably silent
23:49auction for twenty 30-40 years.
23:51But in the setting of a clinical
23:53trial and you only have two,
23:54three and four years.
23:56We felt that a solid endpoint of
24:00Group A versus Group B at four
24:03years would not show a sufficient
24:05treatment effect at that time.
24:06So for us,
24:07slope seemed to be the more
24:08attractive way of doing it.
24:09Next to best we have right now.
24:13You keep up very much, and
24:16if there are any other questions,
24:18maybe they can be directed
24:19to both offline thank you.
24:21Will take the opportunity
24:22to have a small break.