Skip to Main Content

CVM Grand Rounds June 10, 2026

June 10, 2026
ID
14275

Transcript

  • 00:05Everyone,
  • 00:06just wanted to welcome everyone
  • 00:07to our, weekly grand rounds.
  • 00:10Not only the people in
  • 00:11the room, but, across the
  • 00:12state at our watch parties.
  • 00:14Today's a special day, not
  • 00:16only because we are celebrating
  • 00:18the remarkable achievements and,
  • 00:21and career of Doctor. Roosevelt,
  • 00:23which we'll hear about a
  • 00:24little bit in a moment,
  • 00:25but also because we have
  • 00:25a wonderful visitor that's here,
  • 00:28to to to do so,
  • 00:29with Doctor. Chang. So, just
  • 00:31a couple of announcements for
  • 00:33cardiovascular medicine runaround. It's a
  • 00:35CME activity, so please be
  • 00:36sure to get your CME.
  • 00:39We have this is not
  • 00:42we have a few more
  • 00:43speakers coming up.
  • 00:45Maybe I'm not,
  • 00:48I don't know if it's
  • 00:50here.
  • 00:51So just, remind everyone that
  • 00:53next week at this time,
  • 00:55next door,
  • 00:56we'll be running our faculty
  • 00:58translational
  • 00:58research
  • 01:00seminar,
  • 01:01and all are welcome.
  • 01:03I don't know who's speaking
  • 01:04next week. Do you know
  • 01:05Jeff?
  • 01:06Okay. We'll we'll you'll learn
  • 01:08you'll learn about it just
  • 01:09before.
  • 01:11And then, the following week,
  • 01:13and I encourage, hopefully, everyone
  • 01:14to attend,
  • 01:15dear friend, former mentee, and
  • 01:17the current chair of medicine
  • 01:18at Arakan University, Doctor. Zeynep
  • 01:20Samad,
  • 01:21is coming,
  • 01:23fine,
  • 01:24to spend a few days
  • 01:25with us and,
  • 01:26and she is going to
  • 01:27share with her share with
  • 01:29you her, experience
  • 01:31of,
  • 01:31leading an institution,
  • 01:34with and her global health
  • 01:36interests as well as her
  • 01:37interest. She is by training,
  • 01:41a structural interventional a structural
  • 01:43echocardiographer.
  • 01:44And so we'll I don't
  • 01:45know what she's gonna talk
  • 01:46about, but we're excited to
  • 01:47have her here visiting with
  • 01:48us, in a few weeks.
  • 01:49So with that, I'll, ask,
  • 01:52Rachel to come up and
  • 01:54say a few words about
  • 01:55doctor Lowenthal.
  • 01:56And the slides are not
  • 01:57moving, so I'm not sure
  • 01:58exactly why, but
  • 02:02now it's moving.
  • 02:03Oh, so wait.
  • 02:05Do I put up another
  • 02:06slide here? No. Wait.
  • 02:08Because there's this one. Isn't
  • 02:09there that one also? Or
  • 02:11Nope. That's it. What about
  • 02:12this one? Alright. Well Yeah.
  • 02:13This is
  • 02:15Alright.
  • 02:17I'm so pleased to introduce
  • 02:19the Linda Rosenfeld honorary lecturer
  • 02:21today, inaugurated in two thousand
  • 02:23and twenty two in honor
  • 02:24of doctor Rosenfeld's
  • 02:25contributions to Yale cardiology and
  • 02:27to electrophysiology
  • 02:28in general.
  • 02:29Doctor Rosenfeld graduated Tufts and
  • 02:31then, went to Cornell Medical
  • 02:33School where she also did
  • 02:34her internship and residency and
  • 02:36then came to Yale for
  • 02:37a cardiology fellowship in nineteen
  • 02:39seventy nine, and she's never
  • 02:40left.
  • 02:42Doctor Rosenfeld
  • 02:43has mentored every Yale EP
  • 02:45fellow since that time, both
  • 02:47clinically, and I think she's
  • 02:48written a paper with every
  • 02:49fellow in both adult EP
  • 02:51as well as pediatric EP
  • 02:53where she has held a
  • 02:54joint appointment.
  • 02:56Doctor Rosenfeld's dedication to teaching
  • 02:59from making the Goodyear service
  • 03:00what it is today to
  • 03:01the, extensive EKG teaching she
  • 03:04does for our, medical students,
  • 03:06our fellows, our residents,
  • 03:08as well as the,
  • 03:10as as well as the,
  • 03:12fellows clinic, which she is
  • 03:13continuing to mentor in even
  • 03:15into her retirement, has led
  • 03:16to the naming of our
  • 03:17teach of our sectional teaching,
  • 03:19award for, her as well.
  • 03:22For me personally, one of
  • 03:23Linda's, primary contributions
  • 03:26as the first female cardiology
  • 03:28fellow
  • 03:28and faculty member here at
  • 03:30Yale has been to change
  • 03:32the culture in ways that
  • 03:33are pretty dramatic. When Linda
  • 03:33started, it was it was
  • 03:33just fine to pay the
  • 03:34female
  • 03:36When Linda started, it was
  • 03:37it was just fine to
  • 03:38pay the female fellow
  • 03:39less, and the seasonal the
  • 03:40the, social event of the
  • 03:42season was the fellow faculty
  • 03:44baseball game. She's changed that
  • 03:46culture to one of respect
  • 03:47for the contributions of diverse
  • 03:49individuals.
  • 03:50A bit later, at a
  • 03:51time when the university was
  • 03:52not doing right by our
  • 03:53section, Linda's quiet strength galvanized
  • 03:56us to push back. Last
  • 03:57night, Linda told me a
  • 03:59great story from her fellow
  • 04:00days as the, only woman
  • 04:02among eight female fellows.
  • 04:04Apparently, when Linda started, there
  • 04:05were no locker rooms for
  • 04:07the cath lab, and so
  • 04:08the fellows room served that
  • 04:10purpose.
  • 04:11After confirming that Linda had
  • 04:13a brother,
  • 04:14the fellows just went right
  • 04:15ahead and changed into their
  • 04:17scrubs there in the fellows
  • 04:18room. You know, the fellows
  • 04:20sounded like a great group,
  • 04:22but I'm sure there were
  • 04:23many things that were not
  • 04:24easy for Linda at that
  • 04:25time. And, Linda, thank you
  • 04:26for paving the way for
  • 04:27the rest of us.
  • 04:39I'm sorry. Yeah. Yeah.
  • 04:41Next up will be doctor
  • 04:42Yibeng Chung introducing our speaker
  • 04:44for today.
  • 04:50Good afternoon, everyone. Thank you
  • 04:52very much for attending our,
  • 04:55special, seminars. It's our, great
  • 04:57pleasure to welcome doctor Mina
  • 04:59Chang as today's,
  • 05:01speaker for Lindau Rosenfeld honorary
  • 05:03lecture.
  • 05:04And so,
  • 05:06doctor Chang is a professor
  • 05:08of medicine, Cleveland Clinic,
  • 05:10Keshe Western University.
  • 05:12She received her bachelor degree
  • 05:14in, in
  • 05:16in
  • 05:17science from University of California,
  • 05:20San Diego, an MD degree
  • 05:22from University
  • 05:23of Washington,
  • 05:24San Luis. She then pursued
  • 05:27the,
  • 05:28residency,
  • 05:30cardiology fellowship,
  • 05:31clinic cardiology, electrophysial fellowship, and
  • 05:34a poster fellowship,
  • 05:37at WashU.
  • 05:38Doctor Chung is internationally recognized
  • 05:41leader
  • 05:42on atrial fibrillation focusing on
  • 05:44post operative
  • 05:46AF, applying inflammation mechanism,
  • 05:49underlying AF, and investigating
  • 05:51AF, genetics and genomics.
  • 05:54She has funded a Cleveland
  • 05:57Clinic f genomic group with
  • 05:59robust
  • 06:00f for bi reposophy
  • 06:02and a large human atria
  • 06:04tissue and a pie and
  • 06:05a blood biobank.
  • 06:07Very excitingly, doctor Chang has
  • 06:09established,
  • 06:10the robust patient stem cell
  • 06:12program for functional studies of
  • 06:14g o s coupled with
  • 06:15engineered heart tissue
  • 06:17and machine learning and multiarmic
  • 06:20approaches.
  • 06:21She's a center director
  • 06:22for an HRA AI for
  • 06:24research center. She also directs
  • 06:27the researcher center excellence in
  • 06:29cardiovascular
  • 06:30translation of functional genomics.
  • 06:32Her exciting research has led
  • 06:34to impactful publications
  • 06:37in journals such as circulation,
  • 06:38so research,
  • 06:40cell reports, medicine, PNS, data
  • 06:42genets, etcetera.
  • 06:43Doctor Chang's
  • 06:44research has continuously be funded
  • 06:47by major
  • 06:48agencies such as NIH, including
  • 06:50the very
  • 06:51outstanding
  • 06:53program,
  • 06:54program
  • 06:55director project,
  • 06:57so called p o one,
  • 06:58extremely competitive, exciting,
  • 07:00entitled f, poster GOS, mechanism
  • 07:03to treatment.
  • 07:04Doctor Chang has contributed
  • 07:06extensively to research a community
  • 07:09by serving on NIH study
  • 07:10sections.
  • 07:11She is the president of
  • 07:13Heart Rhythm Society
  • 07:15and a chair of the
  • 07:16HIS scientific and clinical
  • 07:19committee, the ECG,
  • 07:21alrysumia committee,
  • 07:22the data science,
  • 07:24and
  • 07:25the COVID rapid
  • 07:27response
  • 07:28research center.
  • 07:30So,
  • 07:31doctor Chung has won many
  • 07:33awards, including
  • 07:35the heart vascular thoracic institute
  • 07:37award, HOA clinic,
  • 07:40clinic,
  • 07:41cardiology council award, HOA
  • 07:43brain science to life award.
  • 07:46Without further ado, let's welcome
  • 07:48doctor Chung
  • 07:50to give her lecture titled,
  • 07:53the insights from atrial fibrillation,
  • 07:56genomics,
  • 07:57from
  • 07:58bedside to bench and, back
  • 08:00again. Thank you, doctor Chung.
  • 08:08Thank you so much for
  • 08:09that very kind,
  • 08:10overstated
  • 08:11introduction.
  • 08:12Thank you very much. I'm
  • 08:13very honored to be here
  • 08:15and honored to to
  • 08:20be able to honor
  • 08:22Linda Rosenfeld.
  • 08:29You'll be able to go
  • 08:30back.
  • 08:31K.
  • 08:34Thank you. We've heard a
  • 08:36lot about Linda already, and,
  • 08:38you know, just she has
  • 08:39been such a pioneering
  • 08:42person,
  • 08:43pioneering
  • 08:44in so many aspects, pioneering
  • 08:46and setting the stage for
  • 08:47all of us women in
  • 08:48cardiology and EP. So thank
  • 08:50you, Linda, again, for doing
  • 08:52that. I mean, I don't
  • 08:53think I would be here
  • 08:54without all the all the
  • 08:56stuff that you took, all
  • 08:57the hardships that you took
  • 08:58for for all of us.
  • 09:01You know, you're known for
  • 09:02your patient care and your
  • 09:04your
  • 09:05managing very complex,
  • 09:07controversial patient challenges. You haven't
  • 09:09shied away from it.
  • 09:10I first got to know
  • 09:11Linda a little more because
  • 09:21about the basics of basic
  • 09:23science behind sarcoidosis.
  • 09:25And she assigned me to
  • 09:26that section. So it made
  • 09:27me, made me look into
  • 09:29it very,
  • 09:30a little deeper. And actually,
  • 09:33from that, I thought, you
  • 09:34know what? I wanna start
  • 09:35study this. And I wanted
  • 09:36to do it in time
  • 09:37for this this, this lecture.
  • 09:39But I have to say
  • 09:40that I haven't completed or
  • 09:42done that yet so I
  • 09:43can't talk about sarcoidosis.
  • 09:44But thank you very much
  • 09:46for spurring that.
  • 09:47And she's done such work
  • 09:49in atrial fibrillation, anticoagulation,
  • 09:51left atrial appendage closure. And,
  • 09:53you know, those are areas
  • 09:54in our field that are
  • 09:55that are really important now,
  • 09:57even controversial,
  • 09:59remote monitoring and, you know,
  • 10:00very difficult issues in congenital
  • 10:02heart disease. She's also I'm
  • 10:04very,
  • 10:05very,
  • 10:07appreciative of your leadership at
  • 10:09Heart Rhythm Society,
  • 10:11and you've been, leading the
  • 10:12patient caregivers committee. And, again,
  • 10:15one of the focuses that
  • 10:16I wanted to bring to
  • 10:17Heart Rhythm Society was to
  • 10:19really strengthen our focus on
  • 10:21patients, and you,
  • 10:22started all that. You were
  • 10:24also on the program committee.
  • 10:25So thank you very much
  • 10:26for all that you've done,
  • 10:27and it's really my great
  • 10:28honor to be here, for
  • 10:30this lecture.
  • 10:32Now what I'm gonna try
  • 10:33to do is to talk
  • 10:35about what we've learned from
  • 10:36AFib genomics and through my
  • 10:38career. You know? We've had
  • 10:39some highs and some lows
  • 10:41and learned from all of
  • 10:42that. But, I'm gonna my
  • 10:44title the title of my
  • 10:45talk is AF Insights from
  • 10:47AF Genomics, going from bedside
  • 10:50to the bench and then
  • 10:51back again and then into
  • 10:53the future.
  • 10:54So we'll we've learned a
  • 10:55lot clinically
  • 10:57from our patients with afib.
  • 10:58And what can we do
  • 11:00to take that,
  • 11:01to spur us to,
  • 11:03address the gaps for basic
  • 11:05translational research? And then how
  • 11:07do we take it back?
  • 11:08So that's what we're gonna
  • 11:09focus about,
  • 11:11on today. So this slide
  • 11:12shows some of the clinical
  • 11:14insights that we can get
  • 11:15from our patients with atrial
  • 11:16fibrillation. As you know, it's
  • 11:18the most common arrhythmia,
  • 11:20sustained arrhythmia. There are over
  • 11:21fifty million people worldwide,
  • 11:23last estimates is that there
  • 11:25are about ten million in
  • 11:26the US. And if you
  • 11:27get to age forty, well,
  • 11:28you have a one in
  • 11:29four chance of getting atrial
  • 11:30fibrillation. It's a very common
  • 11:32arrhythmia,
  • 11:33and it's associated
  • 11:35with stroke
  • 11:36and even higher, like, it
  • 11:38had a,
  • 11:40heart failure and atrial cardiomyopathy.
  • 11:42It was always felt to
  • 11:43be fairly heterogeneous
  • 11:45in terms of its cardiovascular,
  • 11:48associations, associated with a lot
  • 11:50of different types of cardiovascular
  • 11:51stress like obesity,
  • 11:53getting older, heart failure, hypertension,
  • 11:55diabetes, coronary disease, valve disease.
  • 11:58And you could see on
  • 12:00these on these two studies
  • 12:02here, this very marked age
  • 12:04association with AFib.
  • 12:08So in the nineteen fifties
  • 12:10to nineteen nineties, we understood
  • 12:11afib as a reentrant arrhythmia.
  • 12:13There was this multiple wavelet
  • 12:15hypothesis that was put forth
  • 12:17by Gordon Moe and colleagues
  • 12:19where
  • 12:20afib was due to having
  • 12:21multiple
  • 12:22reentrant wavelets in
  • 12:24the atria, the multiple wavelet
  • 12:26hypothesis. So then we knew
  • 12:28that structural substrates were very
  • 12:29important. The larger the atria,
  • 12:31you know, the more likely
  • 12:33you were to sustain these
  • 12:34reentrant circuits.
  • 12:36Then in nineteen ninety eight,
  • 12:37there was this very,
  • 12:39very important paper and paradigm
  • 12:42changing paper that came out
  • 12:43of Bordeaux by Michel Hasegara
  • 12:46and colleagues. And it wouldn't
  • 12:47surprise me if he gets
  • 12:49the Nobel Prize for this
  • 12:50one of these days. I
  • 12:51think a lot of us,
  • 12:52several of us are hoping
  • 12:53and promoting that. But he
  • 12:55published this first paper in
  • 12:57New England Journal of Medicine
  • 12:58where he showed that there's
  • 13:00spontaneous initiation of AFib by
  • 13:02these ectopic beats coming from
  • 13:03the pulmonary veins. And you
  • 13:05can see these little pulmonary
  • 13:06vein potentials that start the
  • 13:08first few beats of atrial
  • 13:10fibrillation.
  • 13:11So that really changed how
  • 13:12we looked at atrial fibrillation.
  • 13:15So we next understood that
  • 13:16AFib could start. The first
  • 13:18few beats could start from
  • 13:19these triggers from the pulmonary
  • 13:21veins. And in susceptible people,
  • 13:24that they could conduct out
  • 13:25to the atrium
  • 13:26and cause the reentrant arrhythmia,
  • 13:29that we know as atrial
  • 13:30fibrillation.
  • 13:32And we also had this
  • 13:33dual substrate model where what
  • 13:36was important were substrates for
  • 13:38AF initiation, those triggers from
  • 13:39the pulmonary veins, and that
  • 13:41you could have substrates for
  • 13:42AF maintenance. So hemodynamic,
  • 13:44the structural basis for sustaining
  • 13:46all these multiple wandering wavelets,
  • 13:48and that you could have
  • 13:49electrical remodeling and structural remodeling
  • 13:51that may, that could occur
  • 13:53and make one more important
  • 13:55than the other move between
  • 13:57this. And there were a
  • 13:58lot of, a lot of,
  • 13:59triggers that could do this,
  • 14:01including stretch, calcium overload, LVH,
  • 14:04ventricular dysfunction, valve disease, autonomic
  • 14:06tone, inflammation, fibrosis, and toxins.
  • 14:09So it seemed to be
  • 14:10such a heterogeneous,
  • 14:13disease.
  • 14:14But we also understood that
  • 14:15the natural history of AF
  • 14:16from the stool substrate model
  • 14:18is that we tend to
  • 14:20that people with AFib tend
  • 14:21to progress. That early in
  • 14:23the in the natural history
  • 14:25of the disease, the sources
  • 14:26for AF initiation, those triggers
  • 14:28may be important. And that
  • 14:29may manifest as just having
  • 14:31frequent atrial ectopy that then
  • 14:33evolves to repetitive runs of
  • 14:35atrial tachycardia, paroxysms of AFib
  • 14:37that become more and more
  • 14:38persistent and even permanent.
  • 14:40And later on, the substrates
  • 14:42for AF maintenance may be
  • 14:44more important.
  • 14:46So with these findings, we
  • 14:48started doing AF ablation. And
  • 14:50the cornerstone of AF ablation
  • 14:51was, found to be pulmonary
  • 14:53vein isolation.
  • 14:54And it turns out this
  • 14:55is a figure of the
  • 14:57Cox Mays procedure. I trained
  • 14:59at, Washington University.
  • 15:01And around that time, Jimmy
  • 15:02Cox was coming up with
  • 15:04this. This was before we
  • 15:05knew about the pulmonary vein
  • 15:06triggers. He had this operation
  • 15:08where his mother He was,
  • 15:11I think very stimulated by
  • 15:12his mother having atrial fib.
  • 15:14So he decided he was
  • 15:15going to work on atrial
  • 15:15fibrillation. He was a very
  • 15:17renowned cardiac arrhythmia surgeon. He
  • 15:19would do AVNRT
  • 15:20and accessory pathways. And what
  • 15:22he did was they had
  • 15:23animal models and, and, then
  • 15:26brought it to humans where
  • 15:27he would make incisions through
  • 15:28the atrium and then,
  • 15:30resew them up. And you'd
  • 15:31have these suture lines. And
  • 15:32the thinking was that the
  • 15:33reentrant circuits would run into
  • 15:34a suture line and you'd
  • 15:36extinguish AFib. But we know
  • 15:37now that one reason why
  • 15:39it was so successful is
  • 15:40that there's a box lesion
  • 15:41around the pulmonary veins.
  • 15:43So this led to a
  • 15:45lot of work on catheter
  • 15:47ablation.
  • 15:48And,
  • 15:48we now we've developed radio
  • 15:50frequency energy methods and cryoablation
  • 15:53to electrically isolate the pulmonary
  • 15:56veins. And then more recently
  • 15:57over the last couple of
  • 15:58years in the US, we've
  • 15:59had,
  • 16:00pulse field ablation,
  • 16:01which has really taken off.
  • 16:03And I remember sitting in
  • 16:04an auditorium like this at
  • 16:06WashU
  • 16:07and,
  • 16:08and we were Jimmy Cox
  • 16:10was doing this live case.
  • 16:11And I remember he said,
  • 16:12I have to tell you
  • 16:14this. This is full of
  • 16:15cardiac electrophysiology. He goes, I
  • 16:16can tell you this. This
  • 16:17is something that those damned
  • 16:18electrophysiologists
  • 16:19are never gonna do with
  • 16:20a catheter. You know? Because,
  • 16:22like, you know, he was
  • 16:23we were taking away his
  • 16:24his business for WPW and
  • 16:26AVNRT, but, anyway so look
  • 16:28at where we are today.
  • 16:31But there are still a
  • 16:32lot of gaps in our
  • 16:33management of AFib. There's a
  • 16:34ceiling of of ablation success.
  • 16:36We can't seem to get
  • 16:37above eighty to ninety percent
  • 16:39success. Even with paroxysmal AFib
  • 16:41and even with post fetal
  • 16:42ablation, we still see issues
  • 16:44with durability.
  • 16:45And we see that there's
  • 16:46recurrence that occur even after
  • 16:48successful PBI. I mean, you
  • 16:50you you if you do
  • 16:51this procedure, you get, you
  • 16:52know, you you you are
  • 16:53converted. You are a believer
  • 16:54of pulmonary vein isolation that
  • 16:56this helps. But then people
  • 16:57can recur, you know, several
  • 16:59years later. And then we
  • 17:00still have pretty poor success
  • 17:02rates in persistent AF. And
  • 17:03and there's a lot of
  • 17:04debate still. Even with post
  • 17:06fetal ablation, what should we
  • 17:07ablate? Should we be fulgurating
  • 17:09the posterior wall? You know?
  • 17:10What should we do beyond
  • 17:11the pulmonary,
  • 17:13vein isolation?
  • 17:14And there's still some risks
  • 17:15with AFib ablation. Even though
  • 17:17it's become a lot safer
  • 17:18with post fetal ablation, we're
  • 17:19not dealing with atrial esophageal
  • 17:21fistula, pulmonary vein stenosis,
  • 17:23those kinds of things. We're
  • 17:24still dealing
  • 17:25with recurrences,
  • 17:27new atrial flutters that are
  • 17:28developing. And,
  • 17:30a late thing that has
  • 17:31been turning up is coronary
  • 17:33spasm that can happen.
  • 17:35And there's a signal for
  • 17:36potential late cardiac
  • 17:39later, not even first early,
  • 17:41within the first few weeks
  • 17:42of sudden death. What's that
  • 17:43all about? We don't know.
  • 17:45Okay?
  • 17:46So, you know,
  • 17:47there is a role still
  • 17:48besides ablation for anturemic drugs
  • 17:51but there are a lot
  • 17:52of challenges with anturemic drugs.
  • 17:53Prorhythmia, intolerance. We haven't had
  • 17:56a new anturemic drug in
  • 17:57over fifteen years. I looked
  • 17:58up Joneterone. It was like
  • 17:59over fifteen years ago. And,
  • 18:02so, you know, back then
  • 18:03we were thinking,
  • 18:04are there any upstream therapies
  • 18:06that can target this AF
  • 18:08substrate?
  • 18:09So our group was looking
  • 18:11at a lot of the
  • 18:11traditional upstream therapies for AFib.
  • 18:13We did a study looking
  • 18:15at c reactive protein and
  • 18:16showed that,
  • 18:17that c reactive protein was
  • 18:19elevated in afib and, you
  • 18:21know, more more and persistent
  • 18:22than paroxysmal.
  • 18:24We did a study out
  • 18:25of the,
  • 18:26CHS study because we saw
  • 18:28that that they had measured
  • 18:29CRP. We were kind of
  • 18:30on the tails of coronary
  • 18:32disease looking at ultra sensitive
  • 18:33CRP. And we saw that
  • 18:35they had also coded for
  • 18:36AFib. And we showed that
  • 18:37CRP could could,
  • 18:39could predict future development of
  • 18:41AFib. So, I was working
  • 18:43with a group of people,
  • 18:44including David Van Wagoner, and
  • 18:45we were looking at, the
  • 18:47influence of upstream
  • 18:50factors like inflammation, calcium overload,
  • 18:52fibrosis, and oxidative stress and
  • 18:54using some of these agents
  • 18:55to try to target it.
  • 18:56So there are a lot
  • 18:57of studies that that came
  • 18:59out of that period where
  • 19:01we were looking at secondary
  • 19:03analyses of randomized trials for
  • 19:04these agents. But, unfortunately,
  • 19:07with the exception of steroids,
  • 19:09which you can't take long
  • 19:10term, and statins for post
  • 19:12op AF that,
  • 19:14and ACE inhibitors and ARBs
  • 19:16in heart failure, the randomized
  • 19:17trials for AF have been
  • 19:19disappointing. You know, it may
  • 19:20work in hypertension
  • 19:22or in, heart failure but
  • 19:23those agents are already, you
  • 19:25know, indicated
  • 19:26there.
  • 19:27So then in the 2000s,
  • 19:29we learned that AF can
  • 19:31be heritable. There are always
  • 19:32some familial cases of AF
  • 19:34around. But we also had
  • 19:34studies coming out of Framingham
  • 19:34that showed that parental AF
  • 19:34increases the risk of
  • 19:36AF studies coming out of
  • 19:36Framingham that showed that parental
  • 19:38AF increases the risk of
  • 19:40AF in offspring. So they
  • 19:42had odds ratios one point
  • 19:43eight five. And then if
  • 19:44you have earlier onset AF,
  • 19:46a parent with AF under
  • 19:47seventy five, the odds ratio
  • 19:49was three. In Iceland, decode
  • 19:51genetics,
  • 19:52also showed that, that AF
  • 19:54increases the risk of AF
  • 19:55in first degree relatives with
  • 19:57about, again, a relative risk
  • 19:58of one point seven seven.
  • 20:00And early onset was,
  • 20:02AF was associated with relative
  • 20:04risk almost five. Patrick Eleanor
  • 20:06at the MGH looked at
  • 20:09lone AF and showed that
  • 20:10that very greatly increases the
  • 20:12risk of AF and family
  • 20:13members. And look at these
  • 20:14relative risks.
  • 20:15Sons, eight. Daughters, nine. Brothers,
  • 20:18seventy.
  • 20:19Sisters, thirty four. And, if
  • 20:21you took a very
  • 20:23detailed history, then you could
  • 20:24find that thirty percent of
  • 20:26people with low AF could
  • 20:27have a family history of
  • 20:28AF. The Danish Twins Study
  • 20:30and the UK Biobank reported
  • 20:32AF heritability to be about
  • 20:33twenty two percent to sixty
  • 20:34two
  • 20:35percent. So perhaps
  • 20:37genetics could be the upstream,
  • 20:39ultimate upstream approach. And around
  • 20:41that time, with the sequencing
  • 20:42of the human genome, we
  • 20:44could assay for markers or
  • 20:45single nucleotide polymorphism,
  • 20:47SNPs, across the genome,
  • 20:50like several hundred thousand of
  • 20:51these markers across the genome.
  • 20:53And the first for the
  • 20:54first time, you can take,
  • 20:55you could get information from
  • 20:57the patient and do these
  • 20:59big, large case control studies
  • 21:01where you take a group
  • 21:02of people with a condition
  • 21:03like afib, a group of
  • 21:04people without afib, and look
  • 21:06at which areas on the
  • 21:07chromosome are associated with your
  • 21:09condition.
  • 21:10And this is like an
  • 21:11unbiased way. Up until this
  • 21:12time, we were looking at
  • 21:14ion channel drugs, electrical
  • 21:16acting drugs, because AF is
  • 21:18an electrical disease. And for
  • 21:20the first time, we could
  • 21:21actually get back from the
  • 21:22bedside from our patients what
  • 21:24genes might be involved with
  • 21:26AFib.
  • 21:27So in the 2000s,
  • 21:29in two thousand and seven,
  • 21:30decode genetics that they had,
  • 21:32genotyped a hundred thousand of
  • 21:34their population with one of
  • 21:35these big, SNP chips. And
  • 21:38they reported that there's there
  • 21:39was an area on chromosome
  • 21:41four that was associated with
  • 21:42afib. Understand that within twenty
  • 21:44four hours, they, you know,
  • 21:45got IRB approval and then
  • 21:47crunched that out.
  • 21:49Meanwhile, at Cleveland Clinic, you
  • 21:51know, we were trying to
  • 21:52follow along the genetics as
  • 21:53well. And we'd started a
  • 21:54lone AF gene bank. And
  • 21:56this is our Manhattan plot
  • 21:57with about almost five hundred
  • 21:59cases for lone AF
  • 22:01in two thousand seven.
  • 22:03So this Manhattan plot on
  • 22:04the x axis is each
  • 22:06chromosome, and each dot is
  • 22:08a marker, the single nucleotide
  • 22:09polymorphism.
  • 22:10And on the y axis
  • 22:11is negative log of the
  • 22:13p value. Because you're looking
  • 22:14at hundreds of thousands of
  • 22:15SNPs, you have to adjust
  • 22:17for multiple testing. So instead
  • 22:18of a p value less
  • 22:20than point o five, which
  • 22:21means five percent of your
  • 22:22results are wrong. Right? You
  • 22:24you need to adjust for
  • 22:25that. So you actually need
  • 22:26a p value of less
  • 22:27than, like, ten to minus
  • 22:28six to ten to minus
  • 22:29eight to be of genome
  • 22:30wide significance.
  • 22:31And you see here that
  • 22:33this is this signal is
  • 22:34actually really strong on chromosome
  • 22:36four. I mean, we we
  • 22:37went into this thinking this
  • 22:38is a heterogeneous
  • 22:40condition. We weren't sure we
  • 22:41were going to see any
  • 22:42kind of genetic
  • 22:43signal in a GWAS.
  • 22:45But we could actually see
  • 22:46this with a hundred cases.
  • 22:48This is like the poster
  • 22:49child of GWAS at the
  • 22:50time, age related macular degeneration,
  • 22:52that they could see a
  • 22:53signal with a hundred patients.
  • 22:55So we actually never,
  • 22:58formally published this paper. We
  • 23:00contributed to the,
  • 23:02to various consortia.
  • 23:03But what's going on at
  • 23:05chromosome four? It's in a
  • 23:06noncoding region. And it turns
  • 23:08out
  • 23:09that the gene nearest to
  • 23:11it is called Pytx2.
  • 23:13Pytx2 is a transcription factor
  • 23:15that's active early in development.
  • 23:17And it is involved with
  • 23:18left right formation of the
  • 23:19heart
  • 23:20and the pulmonary veins. So
  • 23:22we thought, ah, this is
  • 23:23gonna be a piece of
  • 23:24cake. You know? We just
  • 23:25happened to land near a
  • 23:26a gene that's involved with
  • 23:28formation of the pulmonary veins.
  • 23:30So over the years, there
  • 23:31were larger and larger genome
  • 23:33wide association studies. There was
  • 23:35ours,
  • 23:36and then we contributed our
  • 23:37our data to the CHARGE
  • 23:39consortium. It was the NHLBI,
  • 23:42cohorts and heart and aging
  • 23:43research in genomic epidemiology.
  • 23:46And we joined the AFib
  • 23:47working group. So with over
  • 23:48eighteen thousand cases, about ten
  • 23:50years later, there were twenty
  • 23:51four AFG WSLOCI. It's actually
  • 23:53starting to look like Manhattan.
  • 23:54Right?
  • 23:55And then in,
  • 23:57two thousand eighteen, eighty thousand
  • 23:59is looking more crowded.
  • 24:00And this is the latest
  • 24:01from two thousand twenty five
  • 24:03with over a hundred and
  • 24:04eighty thousand cases. And we
  • 24:06have over three hundred,
  • 24:07three hundred and fifty AFGUS
  • 24:09loci. But, you know, these
  • 24:10are really,
  • 24:11small p values. And through
  • 24:13it all, the chromosome four
  • 24:14q twenty five locus is
  • 24:16still the much the very
  • 24:17strongest.
  • 24:18So as I said, this
  • 24:19this gene that's near Pyddix
  • 24:21that is near,
  • 24:24the the locus is Pyddix
  • 24:25two. And it directs left
  • 24:27right
  • 24:29cardiac,
  • 24:30formation, and it's involved with
  • 24:31formation of the pulmonary veins.
  • 24:33It's expressed in the human
  • 24:34and the mouse left atrium
  • 24:36and absent in pretty much
  • 24:37other areas of the heart.
  • 24:39And it suppresses a left
  • 24:40atrial sinus node program.
  • 24:43So maybe there are triggers
  • 24:44here that it is suppressing.
  • 24:46And there's a PidX2 deficient
  • 24:47mouse model, and they're more
  • 24:49susceptible to pacing induced atrial
  • 24:50arrhythmias.
  • 24:52And these mice don't develop
  • 24:53a normal pulmonary vein sleep.
  • 24:55There are a lot of
  • 24:56complex transcription factor interactions, and
  • 24:58it also has a role
  • 25:00in early cardiac regeneration
  • 25:02and repair.
  • 25:03So, you know, could this
  • 25:04really explain pulmonary vein triggers?
  • 25:06As those of you who
  • 25:07do AF ablation, when you
  • 25:08ablate and isolate the pulmonary
  • 25:10veins, you can see sometimes
  • 25:12this isolated firing within the
  • 25:14pulmonary veins that's dissociated from
  • 25:16what's going on. So this
  • 25:18and, and the rest of
  • 25:18the atrium, this is a
  • 25:19slide showing,
  • 25:21a figure showing atrial fibrillation
  • 25:22on the atrium. And within
  • 25:24the pulmonary veins, you see
  • 25:25this firing.
  • 25:26And, and this is, independent
  • 25:28firing during sinus rhythm. It's
  • 25:30hard to make the association.
  • 25:31We had a trend toward
  • 25:32it. But as you know,
  • 25:33you know, especially nowadays, we're
  • 25:35probably ablating a lot of
  • 25:36those, those triggers. So it's
  • 25:37hard to do those associations.
  • 25:40But we did do some
  • 25:41spatial RNA sequence, seeing from
  • 25:43the donor heart tissues to
  • 25:45see whether pit x two
  • 25:47and, other of these transcription
  • 25:49factors are,
  • 25:50expressed in the pulmonary veins.
  • 25:52These are from donor heart
  • 25:53tissues taking the,
  • 25:56taking the the junction or
  • 25:57the pulmonary veins to left
  • 25:59atrium
  • 26:00and,
  • 26:01did,
  • 26:02spatial,
  • 26:03RNA sequencing.
  • 26:04And, we do see pulmonary,
  • 26:07PITUX two expressed in the
  • 26:08pulmonary veins and in the
  • 26:09left atrial appendage. Some of
  • 26:10these are left atrial appendage.
  • 26:12And SHOX two, which is
  • 26:13another,
  • 26:16associated
  • 26:17transcription factor. And HCN four,
  • 26:19which as many of you
  • 26:20may know is the is
  • 26:21the pacemaker current. So we
  • 26:23do see some sporadic,
  • 26:26activity or expression of HTN
  • 26:28four in the pulmonary veins
  • 26:29in these left atrial,
  • 26:30cardiomyocytes.
  • 26:36So this is in a
  • 26:37weird order. But
  • 26:38the question is, can we
  • 26:40use some of these gene
  • 26:41based,
  • 26:43therapies to do, to guide
  • 26:44our intervention? Because I said,
  • 26:45as I said, I thought,
  • 26:47oh, we've hit the jackpot.
  • 26:48We could use this PIVX
  • 26:49to genotype and tell us
  • 26:50who we should do pulmonary
  • 26:51vein isolation on. Well, it
  • 26:53turns out that, I was
  • 26:55wrong. And because these top
  • 26:57loci,
  • 26:58from GWAS,
  • 26:59if anything, are associated with
  • 27:01a higher risk of recurrence.
  • 27:03So with AF recurrence after
  • 27:05cardioversion,
  • 27:06on antrimic drugs, and even
  • 27:08after,
  • 27:08with iblato genetics, some of
  • 27:10these risk variants are associated
  • 27:11with higher risk for,
  • 27:13recurrence. So there's other things
  • 27:15going on.
  • 27:16And it may be that
  • 27:17also there's a role that
  • 27:18I mentioned this role of
  • 27:19PEDX2 in cardiac regeneration.
  • 27:22Jim Martin, who's at Baylor,
  • 27:24has a model where he
  • 27:25looks at been looking at
  • 27:27Piddex two and regeneration.
  • 27:29He has a mouse model
  • 27:30where the neonatal
  • 27:31mice,
  • 27:32their hearts, if you lop
  • 27:33off their eight their apex,
  • 27:35it actually regenerates.
  • 27:36But his pit x two,
  • 27:38deficient mice don't regenerate. He's
  • 27:40also looking that at that
  • 27:41in,
  • 27:42myocardial infarction and trying to
  • 27:45perhaps leverage that. But there's
  • 27:47some he's looking at some
  • 27:48pretty,
  • 27:49complex signaling pathways where it
  • 27:51may may work. But it
  • 27:53made me ask whether or
  • 27:54not this is why we
  • 27:56get recurrences. Because we know
  • 27:59that,
  • 28:00that with atrial fibrillation, we
  • 28:01do see kind of a
  • 28:03change in the transcriptomic
  • 28:04profiling to a more fetal
  • 28:06state. Are we reactivating? Are
  • 28:07we doing something to to
  • 28:09PDX2?
  • 28:10So that's not quite worked
  • 28:11out yet. But maybe it
  • 28:12will tie it together.
  • 28:14Okay. So these are some
  • 28:15of the things that we've
  • 28:16learned from patients on how
  • 28:17do we take it to
  • 28:17the bench. Because once you
  • 28:19found all these things you
  • 28:20know, when you have three
  • 28:20hundred and fifty loci, there's
  • 28:22a lot of work for
  • 28:22basic scientists to work on.
  • 28:24Right?
  • 28:25But it takes a long
  • 28:26time to go from that
  • 28:28SNP to actually function and
  • 28:30actually to treatment.
  • 28:31So what our our labs
  • 28:32have tried to do is
  • 28:33we've tried to follow this
  • 28:34this paradigm that genes to
  • 28:36function going DNA is transcribed
  • 28:38to mRNA,
  • 28:39and then we look at
  • 28:40mRNA, and then we that's
  • 28:41translated to protein, which then
  • 28:43leads to function. And we've
  • 28:44asked a couple of questions.
  • 28:45Like, how is gene expression,
  • 28:47affected by AF itself? And
  • 28:49can we make some inferences
  • 28:51about
  • 28:52gene expression and susceptibility
  • 28:54and and persistence or progression
  • 28:56of AF? And then which
  • 28:58are actually the causative
  • 29:00risk SNPs and risk
  • 29:02genes?
  • 29:02Because the SNPs can be
  • 29:04in linkage disequilibrium
  • 29:05with actually the causal SNPs
  • 29:07or the causal genes and
  • 29:08it may not actually be
  • 29:09next door to it. So
  • 29:12we approached that first,
  • 29:14asking that second question, you
  • 29:15know, what actually are the
  • 29:16causative genes?
  • 29:18So what we had done
  • 29:19over the years in, in
  • 29:20conjunction with David Van Wagner,
  • 29:22in my group, we have
  • 29:23been working very closely with
  • 29:25the cardiac surgeons. And, you
  • 29:26know, those left atrial appendages
  • 29:27that they used to take
  • 29:28off? Well, instead of throwing
  • 29:30them away, we've been hoarding
  • 29:31them to a an atrial
  • 29:33tissue bank. And so what
  • 29:34that's allowed us to do
  • 29:36is to take the, the
  • 29:38genetics and see, and, and
  • 29:40say, and see, look at
  • 29:41which of those snips, those
  • 29:43markers are associated with change
  • 29:45in expression.
  • 29:46So here we have looking
  • 29:47at the nearby, you have,
  • 29:49an allele that's G here
  • 29:50and C and the G
  • 29:51allele is associated with higher
  • 29:52expression. Whereas,
  • 29:53expression, whereas, the c doesn't
  • 29:55make as much. So you
  • 29:57can make inferences about finding
  • 29:59what the causal,
  • 30:00SNP is and the causal
  • 30:01gene is by doing that.
  • 30:02It's called expression quantitative
  • 30:05trait loci. We treat mRNA
  • 30:06expression as a quantitative trait.
  • 30:09So this is another thing
  • 30:10we thought, you know, was
  • 30:11gonna be a piece of
  • 30:12cake. Right? We're gonna find
  • 30:13which of these are causative.
  • 30:15So in this twenty seventeen
  • 30:17Manhattan plot, we had found,
  • 30:20EQTLs,
  • 30:20associations with about half of
  • 30:22them, in blue and light
  • 30:24blue and and red. They
  • 30:26weren't necessarily next to each
  • 30:27other. But, you know, our
  • 30:28top candidate here, we couldn't
  • 30:30see an association with Piddex
  • 30:32too. Darn.
  • 30:33It wasn't quite as a
  • 30:35home run as we thought
  • 30:36it was gonna be. So
  • 30:37why is that? I think
  • 30:38really it's because timing. We're
  • 30:40looking in adults. Okay? And
  • 30:43this PidX2 is active early
  • 30:44in development. So we're kind
  • 30:46of missing that. And then
  • 30:47location, location, location. It's not
  • 30:49just in real estate. We're
  • 30:51looking at left atrial appendage.
  • 30:53But, you know, maybe the
  • 30:54action really is in the
  • 30:55pulmonary vein, left atrial junction.
  • 30:57That's a lot harder to
  • 30:57get the tissue of. But
  • 30:59we did look at it.
  • 31:00I showed you some of
  • 31:00the spatial. We did see
  • 31:02some associations. But it's just
  • 31:03not going to reach genome
  • 31:04wide significance.
  • 31:06So
  • 31:08now we also, we were
  • 31:09also trying to figure out
  • 31:10this age paradox. If we're
  • 31:11seeing all these,
  • 31:13these signals that are active
  • 31:14early in development, how come
  • 31:15people don't get it until
  • 31:16their fifties, sixties, seventies, and
  • 31:18eighties? The age paradox.
  • 31:21So to do that, that's
  • 31:22where adult left atrial appendages,
  • 31:24the tissue may be helpful.
  • 31:27And so what we did
  • 31:28was the first thing we
  • 31:29did was we divided them.
  • 31:30We have about a thousand.
  • 31:31We heavily studied two hundred
  • 31:33and fifty to three hundred
  • 31:34of them. And we divide
  • 31:35them into people who don't
  • 31:37have any history of AFib.
  • 31:38So they're in sinus rhythm,
  • 31:39and then a group that
  • 31:40has a history of AFib,
  • 31:42but they're in sinus rhythm
  • 31:43at the time that we
  • 31:43get the tissue and a
  • 31:44peep and a group of
  • 31:45people that have,
  • 31:47AFib and they're in AFib.
  • 31:49And you can make some
  • 31:50inferences
  • 31:51about AF susceptibility
  • 31:53by comparing the controls, no
  • 31:54AF to the AF sinus
  • 31:55rhythm, and maybe on persistence,
  • 31:57going from paroxysmal to, persistent
  • 32:00from that. The it's very
  • 32:01difficult to interpret though, because
  • 32:03just being an AFib changes
  • 32:04gene expression a a lot.
  • 32:06As you can see with
  • 32:07this heat map, this is
  • 32:08basically those group comparisons, and
  • 32:10these are genes that are
  • 32:11different.
  • 32:12And it's like looking at
  • 32:13alphabet soup, really, to be
  • 32:14honest. It makes you your
  • 32:15eyes kind of glaze over.
  • 32:17It makes my eyes glaze
  • 32:18over. So what you tend
  • 32:20to do with these is
  • 32:21you can do pathway analysis,
  • 32:22and you do what's called
  • 32:23gene set testing, where you
  • 32:24analyze groups of genes because
  • 32:26that can be more robust
  • 32:27than analyzing single genes alone.
  • 32:30So this is another area
  • 32:31in my career where I
  • 32:32thought, you know, we were
  • 32:33we were gonna have it.
  • 32:34I'm going, I'm telling my
  • 32:35group, you know, this is
  • 32:36not gonna be that interesting.
  • 32:37We've been studying inflammation, you
  • 32:39know, and oxidative stress. We're
  • 32:41just gonna see an increase
  • 32:42in that. So to my
  • 32:43surprise, what we saw was
  • 32:45that if susceptibility and persistence,
  • 32:47instead of being,
  • 32:48associated with an elevated level,
  • 32:50was associated with lower levels
  • 32:52of some of these stress
  • 32:53pathways,
  • 32:54a decrease in some of
  • 32:55the transcriptional responses to cell
  • 32:57stress, including some oxidative stress
  • 32:59pathways, protein handling stress,
  • 33:01the heat shock, you know,
  • 33:03factors,
  • 33:04and, that lead to loss
  • 33:06of proteostasis.
  • 33:07And so that was kind
  • 33:10of a new concept for
  • 33:11me, at least. So and
  • 33:12actually focusing on that heat
  • 33:14shock response, a decrease in,
  • 33:16in, proteostasis.
  • 33:18Heat shock response is this
  • 33:20system that's designed to protect
  • 33:22our cells from stress. And
  • 33:23it ensures that your proteins
  • 33:25are folded appropriately. And one
  • 33:27of the hallmarks of aging
  • 33:28is loss of this proteostasis.
  • 33:30And, heat shock factor one,
  • 33:32which popped up in our,
  • 33:34look on other slide, is
  • 33:35the master regulator of this
  • 33:37heat shock response in mammals.
  • 33:38So what happens is when
  • 33:40our cells are exposed to
  • 33:42stress, like,
  • 33:44heat stress, endoplasmic
  • 33:46reticular stress, oxidative stress, the
  • 33:48proteins tend to
  • 33:50unfold.
  • 33:51And cells take care of
  • 33:52that by doing things like
  • 33:54eating it up, autophagy,
  • 33:56and using some of these
  • 33:57heat talk factors to,
  • 33:59to get lysosomal
  • 34:01degradation.
  • 34:02And there's proteasomal
  • 34:03degradation
  • 34:05and then refolding that's chaperone
  • 34:06mediated by some of these
  • 34:07heat shock factors. But when
  • 34:09this pathway is overwhelmed,
  • 34:11then these unfolded proteins,
  • 34:13they precipitate
  • 34:15in aggregates.
  • 34:16And those aggregates, they they
  • 34:18they,
  • 34:19actually precipitate as amyloid.
  • 34:22Okay? And so they're insoluble
  • 34:24protein aggregates from misfolded protein
  • 34:26components.
  • 34:27And,
  • 34:29it turns out that there's
  • 34:30some studies out there that
  • 34:31shows that atrial amyloid is
  • 34:32actually quite common. And there's
  • 34:34one paper. And if you're
  • 34:35over age eighty
  • 34:36ninety, the ninety percent of
  • 34:38those atrial had it had
  • 34:39atrial amyloid. And some studies
  • 34:41have associated, they've done some
  • 34:43co labeling and shown that,
  • 34:46ANP, atri natriuretic peptide, is
  • 34:48associated with some of those
  • 34:49aggregates.
  • 34:50And, you know, you, you're
  • 34:51familiar with ANP. You've probably
  • 34:53talked to some of your
  • 34:54patients who have AFib. They
  • 34:55go into AFib and they
  • 34:56start peeing a lot. The
  • 34:57diuretic
  • 34:58approach. We know that from
  • 34:59our atrial,
  • 35:00appendages, this AMP is really
  • 35:03highly expressed. And so, you
  • 35:05know, this may be one
  • 35:06causative area.
  • 35:08So this is a study
  • 35:09that we did looking at
  • 35:10left atrial amyloid that we're
  • 35:12but we haven't published yet.
  • 35:13But it's a small pilot
  • 35:14study that we're trying to
  • 35:16expand where we did Congo
  • 35:17red staining
  • 35:18in twenty eight left atrial
  • 35:19appendage,
  • 35:20samples
  • 35:21matched for with, with or
  • 35:23without AFib, matched for sex
  • 35:24age and BMI. And you
  • 35:25see that there's an increase
  • 35:27in, left atrial amyloid
  • 35:29associated with AFib, and it
  • 35:30was not associated with age,
  • 35:32sex, hypertension, or BMI.
  • 35:34We also took our,
  • 35:36expression, our our RNA,
  • 35:38seek analysis and did gene
  • 35:40coexpression to see if we
  • 35:41can find other genes that
  • 35:43are associated with not just,
  • 35:44you know, when not just
  • 35:45the GWAS hits. And we
  • 35:46did pathway analysis. And what
  • 35:48we found were a lot
  • 35:50of metabolic pathways that were
  • 35:52implicated,
  • 35:53with oxidative phosphorylation,
  • 35:55mitochondrial dysfunction,
  • 35:56and also the stress response
  • 35:58pathways.
  • 35:59So we devised this dual
  • 36:00risk model, like the perfect
  • 36:02storm for AF, a two
  • 36:03hit hypothesis for AFib, as
  • 36:05you may hear for cancer,
  • 36:08where you may be born
  • 36:09with genetic susceptibility to AFib
  • 36:11that can lead to these
  • 36:12pulmonary vein triggers,
  • 36:14early in life. But it
  • 36:15takes a second hit,
  • 36:17with stressors later in life
  • 36:19before if actually manifests. So
  • 36:21So for example, the stresses
  • 36:22of aging, and I tell
  • 36:23you that is very stressful,
  • 36:25and obesity especially and, can,
  • 36:28can cause a lot of
  • 36:29metabolic protein and mitochondrial stress.
  • 36:32And that may be what
  • 36:34actually leads to AFib. So
  • 36:35in genetically susceptible patients,
  • 36:37inadequate transcriptional
  • 36:39responses to stress may predispose
  • 36:41to AF later in life.
  • 36:43So this gives us a
  • 36:44potential for a lot of
  • 36:45novel upstream
  • 36:46therapies. And, also, it gives
  • 36:48you hope that perhaps you
  • 36:49can actually target some of
  • 36:51these things with lifestyle risk
  • 36:51factor modification and,
  • 36:54defer or avoid AFib.
  • 36:56But what things that we
  • 36:58and others have identified is
  • 36:59that there are potential,
  • 37:01large numbers of novel upstream
  • 37:03targets, candidate pathways
  • 37:05with metabolic stress pathways, mitochondrial
  • 37:07dysfunction, proteostasis,
  • 37:09and a whole host of
  • 37:10candidate genes,
  • 37:12some of whom which are
  • 37:13listed here.
  • 37:15So how do we take
  • 37:16this back to the bedside?
  • 37:17Okay. So can we use
  • 37:19the GWAS data
  • 37:21to help us predict AFib?
  • 37:23Well, this is kind of
  • 37:24a complicated
  • 37:25slide. But it turns out
  • 37:26that each individual SNP in
  • 37:28a GWAS has very small
  • 37:29effect sizes, but you can
  • 37:30combine them to generate these
  • 37:32polygenic risk scores. So this
  • 37:34is a study from Steve
  • 37:35Lubez's group and, that looked
  • 37:38at the predictive accuracy
  • 37:39of clinical versus genetic risk
  • 37:41models for AFib.
  • 37:43And they looked at age
  • 37:44just being older older than
  • 37:45age sixty five. There's a
  • 37:47CHARGE AF score, which is
  • 37:48a clinical risk score. And
  • 37:50then they have this eleven
  • 37:51hundred and sixty eight variant
  • 37:52polygenic risk score and then
  • 37:54a PREDICT AF score that
  • 37:55combined the clinical and the
  • 37:57polygenic risk. And they had
  • 37:59five hundred thousand individuals from
  • 38:00several cohorts. And I think
  • 38:02what, one of the things
  • 38:03we learned is that age
  • 38:05is actually a pretty good
  • 38:06predictor of AF. The AUC,
  • 38:08see statistics here are, like,
  • 38:09point six seven five to
  • 38:11point eight. It's hard to
  • 38:12beat those, actually. And then
  • 38:13if you add the clinical
  • 38:14risk factors, charge AF, it
  • 38:16adds,
  • 38:17a little bit better performance
  • 38:18over age, you know, point
  • 38:20summons, point eight again.
  • 38:22And and if you actually
  • 38:23do the polygenic risk score
  • 38:25alone, it's actually not that
  • 38:26impressive. Right? Point five six,
  • 38:28point six.
  • 38:29If you add it all
  • 38:30together, you do get some
  • 38:31incremental,
  • 38:34improvement in predictive
  • 38:35accuracy. But really, the genetic
  • 38:38polygenic risk score, you get
  • 38:39it's marginally
  • 38:40of value. It's quite modest,
  • 38:42although it is greatest among
  • 38:44younger individuals.
  • 38:46There is a new policy,
  • 38:47some new polygenic risk scores
  • 38:49that includes some that includes
  • 38:50that mega GWAS. So here's
  • 38:52one with one point one
  • 38:54million variants to see if
  • 38:55we can do better.
  • 38:57And it's combined, it's comparing
  • 38:59to, the CHARGE AF clinical
  • 39:00risk score. And again, you
  • 39:02see
  • 39:03c statistics putting them all
  • 39:04together about point eight, point
  • 39:06eight
  • 39:07seven.
  • 39:08So and then when you
  • 39:09add the,
  • 39:10the the polygenic risk score
  • 39:12in these, it adds, you
  • 39:13know, just very small,
  • 39:15modest effect.
  • 39:16So it's kinda hard to
  • 39:18say you can use this,
  • 39:20to try to predict AF.
  • 39:23But I think one thing
  • 39:24that we've learned from all
  • 39:25these genetic studies is that
  • 39:27there's a lot of overlap
  • 39:28of AF with structural heart
  • 39:29disease. Of course, you could
  • 39:30you know, we knew that
  • 39:31already.
  • 39:32But what's really interesting is
  • 39:33if you start looking at
  • 39:34other GWASs like for heart
  • 39:36failure,
  • 39:37there's a lot of commonality
  • 39:38between some of the hits.
  • 39:39So you see this area
  • 39:41on chromosome four near FedEx
  • 39:43two. This happens to be
  • 39:44a very high hit, a
  • 39:45top top hit in this
  • 39:46GWAS of heart failure.
  • 39:48And if you look at
  • 39:49some of the genes where
  • 39:51you look at rare coding
  • 39:52variants that have been implicated
  • 39:53in early onset AI,
  • 39:55then and you you look
  • 39:56at,
  • 39:58genes that have high confidence
  • 40:00for cardiomyopathy
  • 40:01or arrhythmia syndromes in yellow
  • 40:03and, these,
  • 40:05genes that,
  • 40:07are actionable in the middle.
  • 40:09There's a lot of overlap
  • 40:10here. And if you see
  • 40:11this, if you could see
  • 40:12this, a lot of these
  • 40:13are structural genes that you
  • 40:15know from the cardiomyopathy
  • 40:17realm. Okay?
  • 40:19So this is a more
  • 40:21recent study that looked at
  • 40:22sequencing over fifty thousand cases
  • 40:25of AF looking for rare
  • 40:27variants.
  • 40:28Right? As I said, GOS
  • 40:29identifies the common variants that
  • 40:30have very small effect sizes.
  • 40:32But a parallel approach could
  • 40:34be looking for doing the
  • 40:35whole genome sequencing we can
  • 40:36do now in exome sequencing
  • 40:38to identify rarer variants that
  • 40:40actually have large effects.
  • 40:42So this,
  • 40:43this paper from Patrick Eleanor
  • 40:45group, Eleanor's group looked at
  • 40:47sequencing over fifty thousand. And
  • 40:49over half of their population
  • 40:51with AF were diagnosed before
  • 40:53age sixty five. And the
  • 40:54strongest association
  • 40:55was for,
  • 40:57TTN, which codes for site
  • 40:59Titan, a very large protein
  • 41:01with variants that are associated
  • 41:02with cardiomyopathy.
  • 41:04And there are other cardiomyopathy
  • 41:05genes identified too, like myosin
  • 41:07binding protein C3, which is
  • 41:08a hypertrophic cardiomyopathy
  • 41:10gene, lamin A, you know,
  • 41:11implicated in cardiomyopathy
  • 41:13and conduction defects, placophilin
  • 41:15two, which is an ARVD
  • 41:17gene, and some other genes
  • 41:19also associated with arrhythmias. And
  • 41:20there also overlap with GWAS
  • 41:22genes.
  • 41:23So if you look at
  • 41:24sets of these rare variants,
  • 41:25they're associated with, you know,
  • 41:27muscle cell, adipose tissue, cardiac
  • 41:29physiology,
  • 41:30and pathology,
  • 41:31associations,
  • 41:33as well as, you know,
  • 41:34actin, real cytoskeletal,
  • 41:36cardiomyopathy
  • 41:38type,
  • 41:43recommendation that we added to
  • 41:44the AF guidelines a couple
  • 41:46of years back, where there's
  • 41:47a recommendation
  • 41:48from AF guidelines that it's
  • 41:49reasonable to do genetic testing
  • 41:51in patients with early onset
  • 41:53AFib.
  • 41:54And, and these prior studies
  • 41:55from TOPMed and the smaller
  • 41:57studies have shown that seventeen
  • 41:59to twenty four percent of
  • 42:00these may have rare variants
  • 42:01in cardiomyopathy
  • 42:02or channelopathy
  • 42:03genes. A lot of these
  • 42:04genes that we've already touched
  • 42:06upon.
  • 42:07And here's some newer data.
  • 42:09People will say, well, what
  • 42:10are you going to do
  • 42:10with it? You know, we're
  • 42:11going to probably have to
  • 42:12do more,
  • 42:14more surveillance in these people.
  • 42:16So this, is a paper
  • 42:17that comes from Ben Shoemaker's
  • 42:19group at Vanderbilt. And they
  • 42:21looked at two thousand eight
  • 42:22hundred patients that had AF.
  • 42:24And they had genetic testing
  • 42:25done for cardiomyopathy
  • 42:27and arrhythmia genes. And what
  • 42:28they found was that,
  • 42:30two percent had a pathogenic
  • 42:32or likely pathogenic
  • 42:33variant in the titin gene.
  • 42:36And they looked at they
  • 42:37compared, you know, Titan versus
  • 42:38not,
  • 42:40people that didn't have these,
  • 42:42variants. And they found that
  • 42:43the Titan patients more often
  • 42:45had low QRS voltage.
  • 42:47So perhaps when you see
  • 42:48this AF patient and and
  • 42:50they have low voltage considered
  • 42:51genetic testing, and they seem
  • 42:53to have a worse course
  • 42:54with more often having persistent
  • 42:57AF, more often having cardioversions.
  • 43:00And, at the end of
  • 43:01their follow-up, which was about
  • 43:04eight years, I think, I
  • 43:05wrote there, then eleven percent
  • 43:08versus four percent had sustained
  • 43:09VT or VF. And by
  • 43:11then, eighteen percent had ICDs,
  • 43:13forty four percent had developed
  • 43:15LV systolic dysfunction,
  • 43:17and forty seven percent had
  • 43:18a combined endpoint of ventricular,
  • 43:21heart failure, reduced ER, or
  • 43:22sustained VTVR.
  • 43:24So this this genetic testing
  • 43:26couldn't make a difference.
  • 43:29And then,
  • 43:33so bringing this a little
  • 43:34more for,
  • 43:35forward even more, this is
  • 43:37another study that this group
  • 43:38did on impact of management
  • 43:40of doing the genetic testing
  • 43:41in early onset AF. So
  • 43:43they looked at two hundred
  • 43:44forty six patients that had
  • 43:45early onset AFib. They did
  • 43:47genetic testing.
  • 43:48And they found,
  • 43:50a pathogenic or lightly pathogenic
  • 43:52variant in twenty percent. And
  • 43:54they could document a change
  • 43:55in management happened in about
  • 43:57half of them. And so
  • 43:59this, included starting GDMT,
  • 44:01you know, stroke prevention, getting
  • 44:03a new ICD,
  • 44:04or directing lifestyle intervention.
  • 44:07So, in summary, our polygenic
  • 44:09risk scores, with clinical variables
  • 44:11can identify increased risk rate
  • 44:13in the general population that
  • 44:14might then be targeted for
  • 44:16increased surveillance, possibly with more
  • 44:18wearable device monitors. But they
  • 44:20can also potentially identify patients
  • 44:22at risk for AF maybe
  • 44:23after certain more, targeted groups
  • 44:26like cryptogenic
  • 44:27stroke or something like that
  • 44:28that could trigger preventive therapies
  • 44:30like anticoagulation
  • 44:31for people at high risk.
  • 44:33But the genetic testing for
  • 44:34early onset AF is really
  • 44:36here today and can help,
  • 44:38you identify variants in these
  • 44:39cardiomyopathy
  • 44:40genes and perhaps personalize their
  • 44:42arrhythmia management. So, hopefully, in
  • 44:44the future, we'll be able
  • 44:45to use this data to
  • 44:46even identify more novel tar
  • 44:48targets and therapies for individualized
  • 44:50management.
  • 44:52So where do we go
  • 44:53from here?
  • 44:54We were fortunate to get
  • 44:55a program project grant to
  • 44:56try to take forward some
  • 44:58of this information after a
  • 44:59genome wide association study to
  • 45:01look at mechanisms to therapeutics
  • 45:03trying to identify some of
  • 45:04the remaining gaps. So we
  • 45:06have four projects going on.
  • 45:07One is to work on
  • 45:08some of the individual,
  • 45:10GWAS hits and trying to
  • 45:12get them from genes to
  • 45:13function and hopefully new,
  • 45:15drugs. We are looking at
  • 45:17genes and metabolism and looking
  • 45:18at the role of obesity
  • 45:19in some of these mouse
  • 45:20models of AI and, metabolic
  • 45:23stress.
  • 45:24Genes and nutrition. We have,
  • 45:26one looking
  • 45:27a project looking in the
  • 45:28microbiome that looks really interesting.
  • 45:30And then, we have this
  • 45:32genes to omics informed drugs.
  • 45:34So I'm gonna spend the
  • 45:34last few,
  • 45:36minutes talking about, this, genes
  • 45:38to drugs. So it turns
  • 45:40out that genetically
  • 45:41supported targets,
  • 45:42can double the clinical development
  • 45:44success rate. So we're working
  • 45:46with Peisheng Chen who,
  • 45:48uses a network systems biology
  • 45:50approach to trying to repurpose
  • 45:51drugs and for drug discovery.
  • 45:54And what this does is
  • 45:55it takes a lot of
  • 45:56the publicly
  • 45:57available information, protein protein interactums,
  • 46:00and tries to layer a
  • 46:01bunch of other information
  • 46:03on that. So this,
  • 46:05this approach, we're using genomic
  • 46:07data for AF target,
  • 46:10drug target identification. So we're
  • 46:12taking a protein protein interactome,
  • 46:14layering it with the data
  • 46:15that we get from our
  • 46:17tissues with transcriptomics,
  • 46:18our RNA seek, mega GWAS
  • 46:21data, the gen genetic data,
  • 46:22coexpression data, the expression,
  • 46:25and sequence, and protein,
  • 46:27quantitative trait loci, a lot
  • 46:28of the stuff you can
  • 46:29get from from omics, proteomics.
  • 46:31And then we're layering on
  • 46:32single nuclei RNA seek to
  • 46:34try to get more, cardiomyocyte
  • 46:36or other,
  • 46:37other cell based,
  • 46:39specific disease modules. So we
  • 46:41can create an AF disease
  • 46:42module. And then there are
  • 46:43these databases out there that
  • 46:45are drug target databases. And
  • 46:47we can use those drug
  • 46:48target databases to do network
  • 46:50based proximity analysis, trying to
  • 46:51find the closest distance. Like,
  • 46:53trying to find drugs that
  • 46:54act on your genes or
  • 46:56the proteins of interest. You
  • 46:58come up with this candidate
  • 46:59drug list.
  • 47:01So
  • 47:02and then what you can
  • 47:02do is, you know, it's
  • 47:03like bringing Coles to Newcastle
  • 47:05here. You can test these
  • 47:06drugs in these, cardiomyocyte
  • 47:08models that are derived from
  • 47:10inducible pluripotent stem cells with
  • 47:12or without atrial fibrillation.
  • 47:14And these are pretty amazing.
  • 47:15I don't know if you,
  • 47:17see these. When I first
  • 47:18saw this, I was, like,
  • 47:19blown away. It looks like
  • 47:20a map. Right? You know,
  • 47:21there must be some nodal
  • 47:22cells in there, and you
  • 47:23can see the act
  • 47:25activity rippling through the monolayers.
  • 47:27In fact, they're doing it
  • 47:28too much because they're ripping
  • 47:29off the
  • 47:30the media is coming coming
  • 47:32off, so it makes it
  • 47:33really hard to,
  • 47:35to actually measure. But, you
  • 47:36know, it's pretty amazing to
  • 47:38see some of these in
  • 47:39culture, and you can grow
  • 47:40them in engineered heart tissue.
  • 47:42And here, you know, you
  • 47:43have Yibei and other people
  • 47:44who are really looking at
  • 47:46this in a much more
  • 47:47sophisticated way. But this is
  • 47:49one,
  • 47:50one model where you can
  • 47:51grow these within on these
  • 47:52flexible posts and it exercises
  • 47:54them. So that helps them
  • 47:56mature. One of the issues
  • 47:57with, just growing them in
  • 47:58a monolayer that they tend
  • 48:00to be kind of immature.
  • 48:01But you can, you know,
  • 48:02test them, and then you
  • 48:03can do transcriptomic
  • 48:05profiling to show,
  • 48:07that these are more mature,
  • 48:09more like, left atrial appendage,
  • 48:13profiles. You can use these
  • 48:14engineered heart tissues to look
  • 48:16at, the effect of metabolic
  • 48:17stressors. So here you have
  • 48:19one,
  • 48:20in engineered heart tissue, and
  • 48:21then you give isoproterenol, as
  • 48:23you know, it'll make it
  • 48:24beat faster. You can do,
  • 48:25contractility
  • 48:26profiling. You can do optical
  • 48:28mapping.
  • 48:29So we're using this, this
  • 48:31pathway to try to identify
  • 48:33and functionally test drugs, to
  • 48:35target AF progression. So we're
  • 48:37doing that network based, approach
  • 48:39and then testing,
  • 48:41trying to test candidate drugs
  • 48:43in these engineered heart tissues.
  • 48:44And then we'll go to
  • 48:45some mouse models.
  • 48:47I think one of our
  • 48:48issues has been trying to
  • 48:49find a phenotype for AFib
  • 48:51in a in vitro. They
  • 48:53don't fibrillate. You know, your
  • 48:55car your so that's been
  • 48:56difficult. So we're using transcriptomic
  • 48:59profiling at that point. So
  • 49:00this is our first paper
  • 49:02where we, published it. We
  • 49:03were focusing on repurposing FDA
  • 49:05approved drugs because that can
  • 49:06reduce the cost of drug
  • 49:08development. These drugs are already
  • 49:09well studied. So you can
  • 49:10go you can infer,
  • 49:12you know, safety
  • 49:13and effectiveness. So,
  • 49:15this was, the pipeline I
  • 49:17showed you. This is our
  • 49:18first AF disease network, and
  • 49:20it highlights, two hundred and
  • 49:21forty five of these were,
  • 49:24they were,
  • 49:25were AF enriched genes. And
  • 49:26this is our first,
  • 49:28drug target, the top fifty
  • 49:30five drug candidates that were
  • 49:32nominated by the network proximity
  • 49:34analysis. One of the drugs
  • 49:35in here is metformin.
  • 49:37And that really caught our
  • 49:38eye as well because we
  • 49:39were interested in metformin from
  • 49:41one of the, functional,
  • 49:43studies on one of the
  • 49:44hits on FAM thirteen b,
  • 49:46chromosome five. And it turns
  • 49:47out that if he taught
  • 49:48factor one is a master
  • 49:50regulator of protein stress responses,
  • 49:52AMP kinase is a master
  • 49:53regulator of metabolic stress. And
  • 49:55AMP kinase looks at AMP,
  • 49:58ATP ratios.
  • 49:59So it's activated and and
  • 50:01responds,
  • 50:02to increased AMP, ATP ratios,
  • 50:05and it can increase ATP
  • 50:06generation and decrease ATP
  • 50:08expenditures. So this is really,
  • 50:10interesting to us.
  • 50:12And, it turns out that
  • 50:13metformin, a very common prediabetes
  • 50:15and diabetes drug, caloric restriction,
  • 50:18like dieting, weight loss, and
  • 50:20exercise,
  • 50:21have some potential beneficial effects
  • 50:23on AMP kinase. And as
  • 50:25many of you know, there's
  • 50:26our colleagues from, Australia
  • 50:28have done a lot of
  • 50:29studies looking at lifestyle risk
  • 50:30factor modification, weight loss, and
  • 50:32exercise, and showing a decrease
  • 50:33in AFib.
  • 50:35So,
  • 50:36this is the rationale for
  • 50:37trying to look at metformin.
  • 50:39As I've said, you know,
  • 50:40the activity
  • 50:41on AMP kinase, there's also
  • 50:42like a diabetes,
  • 50:44study that seemed to have
  • 50:46lower AFib in people on
  • 50:47in metformin.
  • 50:49We took it to those
  • 50:50iPS cells
  • 50:51and tried to look at
  • 50:52gene expression, and it had
  • 50:53some,
  • 50:55we stressed them with pacing
  • 50:57and saw that there were
  • 50:58some,
  • 50:59favorable effects on heat shock
  • 51:00factor,
  • 51:02expression, protein expression, and also
  • 51:03on some ion channel, calcium,
  • 51:05channel, and and SCN five
  • 51:07a. So we designed this,
  • 51:09study called TRIMAF targeting risk
  • 51:11intervention in metformin for AFib.
  • 51:13And it was a two
  • 51:14by two study,
  • 51:16with forearms control metformin lifestyle
  • 51:19risk factor modification and lifestyle
  • 51:21plus metformin.
  • 51:22And, this was done in
  • 51:23patients that had implanted atrial
  • 51:25devices, pacemakers with defibrillators, so
  • 51:28we could get daily AF
  • 51:29burden and a lot of
  • 51:30activity.
  • 51:31So I'm sad to say
  • 51:33that the first year was
  • 51:34not easy because what happened
  • 51:36was
  • 51:38we hadn't realized this, but
  • 51:39our standard of care group,
  • 51:40we had blasted them with
  • 51:42all of our preventive
  • 51:44cardiology written materials. And we'd
  • 51:46done a lot of modeling
  • 51:47including in that. And if
  • 51:48burden goes up with time,
  • 51:49as soon as they entered
  • 51:51our study,
  • 51:52if burden started going down,
  • 51:53it was hard to to
  • 51:54beat that. And the other
  • 51:56issue was that metformin had
  • 51:57a lot of side effects.
  • 52:00So this was before GLP-1RAs
  • 52:01and that kind of thing.
  • 52:02That was the only thing
  • 52:02we had. Half of the
  • 52:03people had to stop metformin.
  • 52:05So that was difficult, actually.
  • 52:06And I just wonder whether
  • 52:07or not we saw poorer
  • 52:09effects because they had all
  • 52:10these GI side effects. We
  • 52:11probably, they probably had a
  • 52:13higher inflammatory load and all
  • 52:14that. So we're still waiting
  • 52:16on the second year follow-up
  • 52:17and maybe some of that
  • 52:18standard of care effect will
  • 52:19wear off. But, you know,
  • 52:20it made us sit back
  • 52:21and think, well, can we
  • 52:23we what we had done
  • 52:24with it was we had,
  • 52:26saved, all their blood and,
  • 52:29their peripheral blood mononuclear cells
  • 52:31that we can then reprogram
  • 52:32to iPS cells. And what
  • 52:34we could do is is
  • 52:36correlate people who have
  • 52:38a response to metformin or
  • 52:40stress and and and people
  • 52:41who didn't respond and now
  • 52:43probably test it in vitro.
  • 52:45So that's what we're working
  • 52:46on now. And, hopefully, it'll
  • 52:48it could be, like, one
  • 52:49of the first,
  • 52:50to, studies to try to
  • 52:52correlate, you know, response in
  • 52:54a clinical trial, response to
  • 52:55a drug versus in vitro.
  • 52:57So
  • 52:58our next steps are to
  • 52:59try to refine,
  • 53:01the disease network, adding our
  • 53:03single nuclear RNA seq, etcetera,
  • 53:05and improving our prioritization
  • 53:07workflow. A lot of our
  • 53:08our targeting is done using
  • 53:10cancer cell lines for drugs.
  • 53:12And we see on the
  • 53:13list, some toxic drugs. So
  • 53:15to try to narrow it
  • 53:16down, we need to, do
  • 53:17a lot better. So we're
  • 53:18now doing trying to create
  • 53:20a resource where we're looking
  • 53:21at drug targeting onto our,
  • 53:23cardiomyocytes.
  • 53:25So hopefully, we'll have better
  • 53:27news in the future.
  • 53:28But, I I think it's
  • 53:30still a very promising approach
  • 53:32and, look forward to hopefully
  • 53:34updating you in the future.
  • 53:35So in summary, our large
  • 53:36GWAS for AFib implicate over
  • 53:38a hundred, now, you know,
  • 53:39three hundred AF.
  • 53:41The top locus is near
  • 53:42gene implicated in pulmonary vein
  • 53:44formation. And our transcriptomic study
  • 53:46suggests that AF is associated
  • 53:47with impaired or overwhelmed responses
  • 53:49to cell stress. And going
  • 53:51to the bedside, the stool,
  • 53:52the, the bench or dual
  • 53:54risk model proposals that in
  • 53:55genetically susceptible individuals, inadequate transcriptional
  • 53:58responses to stress predisposes to
  • 54:00AF in later life. And
  • 54:02then going from bedside, going
  • 54:03back to that side that
  • 54:04maybe we'll find these dreams
  • 54:06that these drugs that can
  • 54:07target some of these pathways
  • 54:08that,
  • 54:09we can now, study for
  • 54:10AI. And, hopefully, the integration
  • 54:12of all this information will
  • 54:14become easier and easier maybe
  • 54:15in this era of AI
  • 54:16to help us identify new
  • 54:18drugs. So thank you very
  • 54:19much again for inviting me.
  • 54:20Thank you, Linda, for being
  • 54:21such an, inspiration.
  • 54:23And I wanna thank our
  • 54:24team. It takes really a
  • 54:26village to work on all
  • 54:27of this.
  • 54:28And you see a lot
  • 54:29of other names here. Thank
  • 54:31you.
  • 54:34Yeah.
  • 54:37Mina, that was fantastic and
  • 54:39just tremendous amount of work.
  • 54:41Maybe, I'd have to put
  • 54:42Linda on the spot, but
  • 54:43do you have a chance
  • 54:44do you would you like
  • 54:45to ask the first question?
  • 54:47No? Okay. I always ask.
  • 54:50So, maybe I'll start and
  • 54:51then, I'm sure there'll be
  • 54:52others.
  • 54:53You know, you've kind of
  • 54:54projected the opportunity to kind
  • 54:56of pursue
  • 54:57what I'm gonna call facilitated
  • 54:59ablation, maybe, conceptualize it,
  • 55:03of identifying individuals
  • 55:05who might be pursuing,
  • 55:07ablate curative ablation strategies,
  • 55:10who might have high polygenic
  • 55:12risk scores, and then maybe,
  • 55:14possibly pursuing,
  • 55:17therapeutic strategies in the pathways
  • 55:19where they have those that
  • 55:21high risk.
  • 55:22Is that a a construct
  • 55:24that you're pursuing in in
  • 55:25your,
  • 55:26in your program project grant?
  • 55:28And and what do you
  • 55:29think about that idea and
  • 55:30that future?
  • 55:31Well, I mean, that's a
  • 55:32great question because that that
  • 55:34when we looked at those
  • 55:35top GWAS hits and peep
  • 55:37other people have done a
  • 55:38lot of these studies,
  • 55:39they're either no signal Ben
  • 55:41did a a kind of
  • 55:42an ablation
  • 55:44success,
  • 55:45GWAS.
  • 55:46And,
  • 55:47there's a signal that they're
  • 55:48actually associated with higher risk
  • 55:50of recurrence. So I don't
  • 55:51think we can use it
  • 55:52to actually select at this
  • 55:53point. But the question is,
  • 55:55what is that locus doing?
  • 55:57And, why is it that
  • 55:59people are getting it? So
  • 56:00the way we're trying to
  • 56:01approach it is we have
  • 56:02another grant that's been looking
  • 56:03at imaging.
  • 56:04So we have done a
  • 56:05lot of CTs and MRIs.
  • 56:07We're collaborating with Ben Shoemaker
  • 56:09at, Vanderbilt
  • 56:10and, Rod Passman,
  • 56:12with the MRI at Northwestern
  • 56:14and Anat Madhubushi, who used
  • 56:16to be at Case, now
  • 56:17at Emory,
  • 56:18doing a lot of radiomic,
  • 56:20machine learning,
  • 56:22analyses of the images to
  • 56:24see what's different
  • 56:26genetically
  • 56:27about the pulmonary veins. So
  • 56:28it's taken us a long
  • 56:29time because they try to,
  • 56:33associate with a pulmonary vein
  • 56:35morphology has been really challenging.
  • 56:36But that's something that we're
  • 56:37looking at. And are there
  • 56:39areas
  • 56:40of of the atria that
  • 56:41we should be targeting more?
  • 56:42And I have learned from
  • 56:43it. I mean, there's some
  • 56:44areas that we've seen that
  • 56:46are associated with recurrence, you
  • 56:47know, like inferior to the
  • 56:49right things like that that
  • 56:51that already changing my
  • 56:53my my practice without, you
  • 56:55know, without,
  • 56:57you know, major publication. But
  • 56:59I think that's an area
  • 56:59that we really wanna
  • 57:01hit.
  • 57:05Mina,
  • 57:06wonderful talk. I'm it it's
  • 57:08amazing that you propose this
  • 57:10to hit a a mechanism
  • 57:12for f. So I'm wondering
  • 57:14besides the pulmonary vein trigger,
  • 57:17you mentioned in f,
  • 57:18patient, they have reduced
  • 57:20stress response pathway.
  • 57:23What is the potential
  • 57:24reasoning for those f patient?
  • 57:26They have reduced stress response.
  • 57:29Yeah. It's kind of hard
  • 57:30to say. I mean, is
  • 57:31it AF itself that is
  • 57:33overwhelming,
  • 57:34that we're that that these
  • 57:36things are overwhelming? So the
  • 57:37the the reason why it's
  • 57:39decreased is not that
  • 57:41not that known. And, you
  • 57:42know, I think we were
  • 57:43talking yesterday, today, where in
  • 57:45paroxysmal may some of these
  • 57:46pathways are increased, but once
  • 57:48you get to persistent,
  • 57:49they're decreased. So what's going
  • 57:50on there? Are we overwhelming
  • 57:52overwhelming it? You know? And
  • 57:54I think that's really a
  • 57:55a question to be answered.
  • 57:57And there's there's a lot
  • 57:58more complexity there that we
  • 57:59have not even begun to
  • 58:00touch because on that one
  • 58:02slide, there's actually,
  • 58:04a signal for increased
  • 58:06myocyte differentiation.
  • 58:07So what's going on there?
  • 58:08You know, are we
  • 58:09going to cardiofibroblast,
  • 58:11fibocardium, or whatever those things
  • 58:12are? Are we so so
  • 58:14there's a lot more there
  • 58:16to unpack, but it's a
  • 58:17great question in which I
  • 58:18knew the answer to that.
  • 58:22Thank you so much for
  • 58:22an amazing talk and amazing
  • 58:24research.
  • 58:25My question is about the
  • 58:26GWASes taking covariates
  • 58:28for the environmental
  • 58:30risk factors
  • 58:32and, what your thoughts are
  • 58:33about that if those if
  • 58:34the studies did take things
  • 58:36like obesity, BMI,
  • 58:38alcohol consumption as as covariates,
  • 58:40and whether we should be
  • 58:41doing that or whether we
  • 58:42should actually be looking at
  • 58:43gene environment interaction to kinda
  • 58:46maximize that signal rather.
  • 58:48There used to be more
  • 58:49done with that. We haven't
  • 58:51really
  • 58:51you know, because that would
  • 58:52be really important for us
  • 58:54to see what how how
  • 58:56modifiable
  • 58:57this really is. And so
  • 58:58that's been a challenge incorporating
  • 59:00all that. And and a
  • 59:01lot of these have been
  • 59:02done in big population studies.
  • 59:04So I think and and
  • 59:05some, that kind of data
  • 59:07are are,
  • 59:09are, you know, collected. So
  • 59:10I think within charge, they
  • 59:12probably have that.
  • 59:14I don't know that we
  • 59:14actually incorporated that in any
  • 59:16of these big g mega
  • 59:17GWASes, but maybe that's the
  • 59:19next step. It's a great
  • 59:20idea.
  • 59:22A a great talk. One
  • 59:23of the benefit of genetic
  • 59:25testing is that actually,
  • 59:27you can do segregation
  • 59:28analysis and do cascade screening.
  • 59:31But the problem I see
  • 59:32here is that inheritance from
  • 59:34the father to child is
  • 59:35much lower than sibling.
  • 59:37One reason being actually that
  • 59:38your SNPs are actually randomly
  • 59:41distributed in among your children.
  • 59:43So we cannot use that
  • 59:44for
  • 59:45cascade screening and, you know,
  • 59:48understanding who is at risk
  • 59:49at from the children of
  • 59:50AFib. We know if a
  • 59:52person has AFib has AFib
  • 59:53already. But the children, we
  • 59:55cannot do that. Now you
  • 59:56have also rare variants, but
  • 59:58they also have a very
  • 59:59low penetrance, for example, titin.
  • 01:00:01So this is how you
  • 01:00:03wanna answer that. Second thing
  • 01:00:04I wanna have actually at
  • 01:00:06the basic level. You mentioned
  • 01:00:07actually amyloid,
  • 01:00:09kind of deposition,
  • 01:00:10autophagy, and then you came
  • 01:00:12AMPK that actually if it's
  • 01:00:13defective causes
  • 01:00:14reduced autophagy. I wonder if
  • 01:00:16that could be a readout
  • 01:00:17for your
  • 01:00:18single for for your iPSCs
  • 01:00:20because you don't have actually
  • 01:00:21great readout. That actually maybe
  • 01:00:23autophagy is actually the one
  • 01:00:24of the major
  • 01:00:25readout for the drug.
  • 01:00:27Yeah. Those are great questions
  • 01:00:28and suggestions. So I think,
  • 01:00:30you know, this idea of
  • 01:00:31penetrance is really interesting. To
  • 01:00:33be able to take your
  • 01:00:35familial,
  • 01:00:37you know, your familial disease
  • 01:00:39and you know what their
  • 01:00:40genotype is, and then you
  • 01:00:41have a family that, you
  • 01:00:43know, maybe you have penetrance
  • 01:00:44or not. And then maybe
  • 01:00:46incorporate the GWAS, the the
  • 01:00:48polygenic risk scores, the the
  • 01:00:51common variants and how they
  • 01:00:52influence penetrance would be really
  • 01:00:54interesting.
  • 01:00:55I don't know that we
  • 01:00:56have done that in AF
  • 01:00:58as of yet, but that's
  • 01:00:59would be one approach, I
  • 01:01:00think.
  • 01:01:01And then
  • 01:01:02you're right. For the for
  • 01:01:05we don't have a good
  • 01:01:06phenotype
  • 01:01:07in an iPS cell cardiomyocyte.
  • 01:01:09They just, you know, they
  • 01:01:10don't
  • 01:01:11they don't even have a
  • 01:01:12lot of
  • 01:01:14a lot of irregularity. I
  • 01:01:15think we can look at
  • 01:01:16calcium,
  • 01:01:17studies. It's not high throughput.
  • 01:01:19But if we can find
  • 01:01:20a,
  • 01:01:22you know, like a signature
  • 01:01:23of AF that we can
  • 01:01:24then assay for biochemically
  • 01:01:26or with, you know, RTPC
  • 01:01:28or something, that would be
  • 01:01:29that's what we're working on
  • 01:01:30next because
  • 01:01:31and I and I have
  • 01:01:32to say over the last
  • 01:01:33year, I probably should have
  • 01:01:34come on that a lot
  • 01:01:35earlier and worked on that,
  • 01:01:37earlier in this grant to
  • 01:01:38try to figure out we
  • 01:01:40were we that we were
  • 01:01:41we've been working so hard
  • 01:01:42on trying to figure out
  • 01:01:43the right triggers and stress
  • 01:01:45to do, but still, you
  • 01:01:46know, the phenotype is really
  • 01:01:48difficult if you're relying on,
  • 01:01:50RNA seek to what we
  • 01:01:52were what we had proposed
  • 01:01:53that we were gonna use
  • 01:01:54an RNA seek profile where
  • 01:01:56we have our left atrial
  • 01:01:57appendages in AF versus non
  • 01:02:00AF in sinus rhythm. So
  • 01:02:01you have a transcriptomic
  • 01:02:03profile there. And the goal
  • 01:02:04is with pacing stress. I
  • 01:02:06didn't show you some of
  • 01:02:07the stuff we have done
  • 01:02:08with pacing stress and, and
  • 01:02:10testing the drugs.
  • 01:02:12We are doing pacing stress,
  • 01:02:14and we look at the
  • 01:02:15the gene's expression
  • 01:02:17that is going in a
  • 01:02:18similar direction as AFib might
  • 01:02:20in the left atrial appendage
  • 01:02:21in the human, and then,
  • 01:02:23looking for drugs that then
  • 01:02:25reduce it back to a
  • 01:02:26more sinus rhythm profile.
  • 01:02:28So using that as a
  • 01:02:29potential score is some way
  • 01:02:32to to try to narrow
  • 01:02:33the list. Otherwise, you're with,
  • 01:02:34like, that heat map. You
  • 01:02:36have this
  • 01:02:37dizzying
  • 01:02:38list of drugs, and how
  • 01:02:39are you going to,
  • 01:02:41how are you going to
  • 01:02:42curate them other than just
  • 01:02:43manually? So but you're you're
  • 01:02:45right. Maybe looking at some
  • 01:02:47mitophagy
  • 01:02:48type,
  • 01:02:49marker might be actually
  • 01:02:51helpful too. Yeah. Thanks. Glad
  • 01:02:53to be, thanks, doctor Tom,
  • 01:02:55for coming in.
  • 01:02:57Thank you. Thank you so
  • 01:02:59much.