My name is Tommy Dewland
and I'm going to cardiac electrophysiologist
here at UCSF. It is my distinct pleasure to introduce my colleague,
Dr. Greg Marcus. Dr. Marcus is a Professor
of Medicine and is the associate chief of
cardiology for research. He truly is a world-renowned
clinical investigator and has published nearly
300 peer review papers. In addition, Dr. Marcus is a highly skilled and experienced
cardiac electrophysiologist. Now, as students of medicine, I am sure you have seen or will shortly appreciate
that in healthcare, we often use seemingly
complex jargon to describe fairly simple thing. In this case, a cardiac electrophysiologist is a heart rhythm specialist. We are in essence
electricians of the heart. To put this career
pathway into perspective, it involves four years
of medical school, three years of
internal medicine, three years of
general cardiology, and then two years specifically devoted to the diagnosis and treatment of heart
rhythm abnormalities. With that, I wanted to hand
things over to Dr.
Marcus who will be speaking to us this evening about atrial
fibrillation. Thank you so much for
the kind introduction, Tommy. Welcome, everyone. We're very excited
to have you all here and hope to provide some interesting and
hopefully useful information and get started on this talk, which is going to focus
on atrial fibrillation. This was largely motivated by the fact there's been
so many advances in atrial fibrillation. I've given a talk in
this setting before, which as many of you may know, leads to a presentation on YouTube and many of
my patients Tommy, they've seen some of
these talks and I've become very aware
that my last talk, which on this subject
in this forum, which I believe was in 2015
is now willfully out of date. Also, we and others
have contributed novel findings
regarding the roles of alcohol, or caffeine, of course, Apple Watches, and their relationship to atrial fibrillation as a new development
pertinent to many.
We thought this
would be hopefully, again useful and
interesting talk. Oftentimes people
gloss right over their disclosure
slides but I think given the nature of this direct interaction
with the public, I think it's worth spending
a little bit of time on these disclosures so
that these are clear. I've received research
from the NIH from a not-for-profit called the Patient-Centered Outcomes
Research Institute. Another taxpayer-funded
entity focused on tobacco-related research
in California, that's TRDRP. I have received funding from a private for-profit
company called Baylis, they make tools
specifically related to generally one part of the
catheter ablation procedure where we cross from the right
atrium to the left atrium. I'll touch on that briefly. I've been a consultant for a startup company called
InCarda therapeutics. They make an inhaled
therapeutic in the hopes of essentially delivering
drugs directly to the heart through the lung to convert atrial fibrillation. I'm also a consultant
for Johnson and Johnson. That's related to work on a steering committee related
to a randomized trial. I also do hold stock
InCarda Therapeutics as in fact helped to
found that company. I'm not going to
be talking about that technology
or that approach, although happy to answer
questions if that's of interest about it, given time.
Perhaps especially important and especially related
to what will end up being towards the
end of the talk. I have not received any funding from the food and
beverage industry, which I know can be an issue when it comes
to those things. I certainly have no
conflicts there. Jumping right into this common rhythm disturbance called atrial fibrillation, what is, so, in order to
understand what it is, it's really important to
first understand what the normal electrical
condition of the heart is, or normal sinus rhythm, shown here on the left, which originates in a structure in the right upper chamber, the right atrium
called the sinus node, that fires about 60
times per minute, but it can fire 80
times per minute, 90 times a minute. It is sensitive to adrenaline, so it goes up with
more adrenaline, which can happen with exercise or excitement or anxiety, etc.
Then that electrical
signal propagates through the muscle cells
of the top two chambers, the right atrium and
the left atrium. They're all
electrically connected. However, the atria
are not usually directly electrically
connected to the lower chambers,
the ventricles. Here we have the right ventricle
and the left ventricle. The heart valves,
two of them are in the way and they do
not conduct electricity. In the great majority of cases, there's a rare circumstance
I'll mention in a second. Then the great majority
of circumstances, the only way for that
electrical impulse to get from the atrium to the ventricles is through this structure
called the AV node. That then leads to what's
called the bundle of His, which is a specialized
conduction system that then breaks into the left bundle
and the right bundle. Those then lead to these
Purkinje fibers that permeate the ventricles and electrically
activate the ventricles, which subsequently leads to
their muscular contraction. The one rare case that we also treat as
electrophysiologists, where the atria are directly connected to the
ventricle is when there's a little thin piece of
muscle that is just leftover from development in the womb and that can lead to
other arrhythmias.
That can lead to something or is associated with something called the Wolff Parkinson
White Syndrome, which will not be a focus
of the current talk. But again, happy to answer
questions about that if people are curious. In contrast to this, we consider the condition of atrial fibrillation
where the top chambers are very rapidly
contracting in a chaotic, unpredictable fashion, but the AV
node, the His-Purkinje system, the ventricles are still
intact and they are responding as I will describe
to the atrial fibrillation. In terms of the general
outline for this talk, I will focus on a little
bit of epidemiology, then talk about how we make a diagnosis of atrial
fibrillation that will be pertinent to the role of the Apple Watch and its
technologies or capabilities, as well as some
other smartwatches. Then what are the consequences
of atrial fibrillation? Then a bit about
how do we treat it, and then how do we prevent it, which is a new way to think
about atrial fibrillation, and which is
especially pertinent, it turns out, to lifestyle
factors and this is where I will discuss the role
of alcohol and caffeine.
Now, as we think about this
common rhythm disturbance, it's important to establish some definitions and really
a key feature of it, which is that it can
be intermittent. We call that paroxysmal
atrial fibrillation or it can be consistent
or persistent, which is the formal name. Previously, what we now refer to as persistent
atrial fibrillation, meaning it is just
going to continue unless we actively do
something to stop it, was also called chronic
atrial fibrillation.
There's another category called permanent
atrial fibrillation, which as implied means
that atrial fibrillation is just going to
continue on forever. Really, that mainly occurs
when there's a decision made, ideally in partnership between the patient and the treating
physician that we're just going to allow the
atria fibrillation to persist without any
intended interruption. Now importantly, the
same patient can have paroxysmal and persistent
atrial fibrillation at different times. They may have an episode
that is persistent and we shock them out of it
back to a normal rhythm, then they have a pyrrhiccism
that lasts an hour. Although patients tend to be either paroxysmal
or persistent, most of the time, certainly the same patient can
experience both conditions. In terms of the epidemiology, the main point is that
atrial fibrillation is very common. Certainly more than five
million cases, probably now, around 10 million in the US alone expected to be 12 million, probably more by 2030. Again, the lower end of the estimate is that there are 50 million cases
around the world.
Once a person hits
the age of 40, the lifetime risk of
subsequently developing atrial fibrillation
is one in four. What are the risk
factors for this? Age is probably the most
potent risk factor. As we grow older, we all are at heightened risk
for atrial fibrillation. If you have a room where everyone is at least
60 years of age, about five percent, if not more of them have
atrial fibrillation, a room filled with
octogenarians and older, at least 10 percent of them will have
atrial fibrillation. Family history of
atrial fibrillation is an important risk factor, but it's not a classic
Mendelian inheritance, so it's not that,
well, my dad had it, therefore, I have a 50
percent chance of getting it. It's not that simple. It's much more
varied and complex than that and actually,
it depends on the family. There are many genes that
may be responsible for atrial fibrillation and
some families have been described where it
really is Mendelian, it really is a 50
percent chance of getting a fib if
one parent has it.
In most cases, it's
what we call polygenic, meaning due to a combination
of probably many, many common genetic variants. It's not that someone is
absolutely destined to get atrial fibrillation
or destined to never experienced
atrial fibrillation, but rather dials up or
down the susceptibility. There's almost certainly
environmental factors that then interact with that propensity to render one more or less prone
to atrial fibrillation. Other common risk factors
include European ancestry. Dr. Dolan and I have published fairly extensively
on that subject.
Those with higher blood
pressure, diabetes, heart failure, coronary
artery disease, including a history
of a heart attack. Those are the
obstructive sleep apnea. Those all increase the risk, and then we're recognizing
more and more that there are these more readily
modifiable risk factors such as obesity and alcohol use. We'll dive more into that later. Then there are these
special circumstances. There are some conditions where atrial fibrillation
might be reversible. Hyperthyroidism. When the
thyroid gland is overactive, certainly that's associated with a heightened risk of
atrial fibrillation.
There's some evidence,
although frankly, it's not super robust, that once that
hyperthyroidism is treated, that the atrial
fibrillation may resolve. Post open heart surgery
really substantially increases the risk for
atrial fibrillation. Of everyone undergoing
open heart surgery without any history
ventricle fibrillation, 20-30 percent of
them will exhibit, will develop atrial fibrillation
during their recovery. Usually occurs about
three to four days later.
The thinking is the
evidence suggests that in most of those cases
it actually will resolve. We still don't fully understand their long-term risks
of atrial fibrillation. Then pericarditis, which is inflammation of
the heart lining, which I think folks have
become more aware of it because of
myocarditis related to COVID and COVID vaccines. Pericarditis is related to that. Similarly may represent a reversible cause of
atrial fibrillation. Now, also, people can
certainly develop atrial fibrillation
in the absence of all of these risk factors. That represents about 30
percent of all cases. Some people, young, none of the
characteristics I just described can yet still
develop atrial fibrillation. This is sometimes called
lone atrial fibrillation. There is pretty good evidence that those individuals probably have more of a
genetic propensity to developing the disease. So now we will move into the consequences of
this common arrhythmia. This is a general outline
for the consequences. Then I will dig in
a little bit more, especially on bullets 1 and 3. We worry about reduced
quality of life.
People can be quite symptomatic. Adverse remodeling of the heart. Usually this is attributed to prolonged fast
ventricular rates or a fast pulse that can lead
to weakening of the heart. There's some more recent
evidence that even in the setting of a
normal heart rate, such as when atrial fibrillation is treated with medicines, that the heart may still adversely remodeled by
adverse remodeling, I mean, can become larger,
can become weaker, can be more prone
to heart failure, more prone to leaky
valves as that part gets larger and the valves are unable to close as well
as they otherwise would. Then very important, we worry about something
called thromboembolism. Thrombo refers to the formation of a thrombus or a blood clot. Then embolism means
anything that is traveling in the bloodstream
from one place to the other.
What can happen is these clots
can travel from the heart. They're formed in the setting
of atrial fibrillation and travel to some blood
vessel, block it, occlude blood flow and oxygen, starving that tissue that receives that blood
flow of oxygen, leading to cell death. The most feared complication
here is stroke. If we put a catheter
in the atrium, one of the top chambers when someone is in atrial
fibrillation, the rate is incredibly fast. This little strip
here is actually from a pacemaker lead that's
sitting in an atrium, in a patient with
atrial fibrillation. Oftentimes you can count these going more than
400 beats per minute, 500, 600 beats per minute. If the ventricles go that fast, that is not compatible with life unless maybe
you're a hummingbird, but certainly not in humans. One of the important
consequences of going so fast in the atria is you
lose the normal blood flow, the usual atrial kick, and therefore blood stagnates.
Whenever blood sits
still in one place, it is prone to forming
a thrombus or a clot. There's one part of the
atria that's especially vulnerable to this and that's this structure called the
left atrial appendage. It is literally an appendage that comes off the left atrium. This is what's called
a transesophageal echocardiogram with a view of
the left atrial appendage.
The esophagus is the
swallowing tube, which is in the back part of
the chest behind the heart. The left atrium
actually is a little bit behind the right atrium. The right atrium is towards
the front of the chest, the left atrium
towards the back. We don't get a great view of the left atrium when we put an ultrasound probe
right on the chest wall. But if we have the
patient swallow that probe or under
advanced that probe, we can get a very clear, beautiful view of
the left atrium and especially the
left atrial appendage. This is showing blood clots in the left atrial appendage and a patient with
atrial fibrillation. It's important to
mention a concept because this is what
I just described, is the classic reasoning used to explain why
atrial fibrillation, which leads to
stagnation of blood flow because of the rapid rate, then leads to blood clots. This then leads to a common question that patients
asked me all the time. Very understandable
and very intuitive.
There's a somewhat
counter-intuitive answer to that question, which is why it's
now important to skip ahead a little
bit to treatments, which I will come back to. That is a common question, is okay, I see what
you're saying. Afib lead to blood clot, that's
what leads to the stroke. If we don't allow my
atria to defibrillate, we keep my heart
in normal rhythm, that should be enough
to prevent those clots. But it actually doesn't, at least in many patients. It just hasn't been
shown to be true. I'll talk a bit more about that. When it comes to treatments, the thing that prevents the
clots are blood thinners. Now, again, that's
counter-intuitive given what I just said.
There's perhaps a useful
way to think about this, which I believe is almost
certainly operative, at least in some patients. I can't say all patients. But it may be that there are some atria that are
prone to forming blood clots in the
left atrial appendage because there's some scar
tissue there, for example. those same atria are also
prone to fibrillating.
In those cases, the atrial
fibrillation may be more of a epi phenomenon or a
marker to say, "Hey, this is an atria that's prone to forming a blood clot,"
and in that circumstance, getting rid of the afib, suppressing the afib
is not going to be sufficient to prevent
the blood clot. I've seen this in
some of my patients. I think the reality is there's probably a spectrum
and that's the case. In some cases, there are
other cases where it clearly really is directly
related to the afib. The other issue is
that, as I described, the Afib itself lead to
remodeling and, for example, enlargement of the atria
and may itself over time change the characteristics
of the atria to make them more prone to clotting.
This is clearly not a
straightforward issue, but very important
to appreciate. Now I talked about that AV node. One of the key
characteristics of the AV node is that it just
can't conduct that fast. It cannot conduct more
than 400 beats per minute. This may be something that
we evolved to prevent. Otherwise, humans would die when they had
atrial fibrillation. The AV node, in this case, really saves us, in that it generally
will conduct maybe 100, 110, 120, 130 beats per minute
during atrial fibrillation, certainly very much
compatible with life. Although perhaps the
ventricles will go fast enough to make the
person not feel well. But atrial fibrillation,
importantly, is not generally an
imminently dangerous rhythm. It's not a true emergency, largely thanks to this AV node. Now the AV node is heavily influenced by what we
call autonomic tone, which is this tension
or balance or harmony between the
adrenaline side of the nervous system, which is called the
sympathetic nervous system, versus the vagal side, which is also called the
parasympathetic nervous system, which slows things down.
If one had a lot of vagal tone, such as their resting, they just had a big meal, the AV node will be
relatively slow. That person may be in
atrial fibrillation with a ventricular rate, which will result in the pulse of about 80 beats per minute. Then they run up the stairs, they hear the phone ringing in. The phone happens
to be upstairs.
They run up the stairs, they get more adrenaline
and suddenly, whereas in normal sinus rhythm, their heart rate
might go 110, 120, now it's going 160, 170 due to the afib. Now, the ventricles
which generate the pulse will not only
be faster than usual, they'll also beat irregularly. This irregularly
irregular pattern, it really is random, is very characteristic
of atrial fibrillation. Another consequence
of all of that is that this combination of
loss of the atrial kick, as well as the ventricles beating a bit fast as well
as beating irregularly, that all leads to less ventricular filling and a reduction in ventricular
output or cardiac output, and all of that can lead
to various symptoms, including fatigue,
shortness of breath, chest discomfort, palpitations, just feeling faint and sometimes
just not feeling well. Interestingly, some people
are completely asymptomatic, and we still don't completely
understand why that is. Then even the people
who swear they know when they're in atrial fibrillation because
they have symptoms, if you put monitors on
them in many cases you will find there having
asymptomatic episodes. In terms of the diagnosis, we use the electrocardiogram.
There's a picture here,
the electrocardiogram. Very common test. Very quick. It's when they put stickers or electrodes on your chest
that they connect to wires. A normal sinus rhythm is
shown here on the top, we're looking for
these little waves that are marked by
these asterixis, which we call the P wave
and that represents the organized conduction
of the atria. Then with the solid
arrows, these larger, more rapid waves are
called the QRS complexes. Those represent conduction
of the ventricles. They're rapid because this his bundle branch
Purkinje network conducts very rapidly and you'll notice these QRS
complexes are much bigger than the P waves because the ventricles generally
have a lot more meat, so they're much more substantial than the relatively
smaller atria.
Of note, this wave marked by this dashed arrow
is called the T wave, and represents the
electrical repolarization of the ventricles, which then might lead
to the question, well, what about electrical
repolarization of the atria? It is there, but it's
very hard to see. Again, partly because the
atria are much smaller and probably is buried
largely in the QRS. Now, in contrast to this, we look at an electrocardiogram
in the setting of atrial fibrillation where we will have an absence
of these P waves. The underlying baseline
is somewhat undulating. Now you may notice the
T waves and think, "Oh wait, how do you know
those aren't P waves?" But there are broader
than the P wave. They reliably follow
the QRS complexes, and then the QRS complexes, again representing
ventricular contraction, are more rapid and they
are irregularly irregular. This electrocardiogram,
which we also refer to as an ECG or from the German EKG, is the gold standard
really the only way to definitively make a diagnosis
of atrial fibrillation.
This brings us to smartwatches and some of the new
technologies that are designed specifically
to identify individuals with
atrial fibrillation. To talk about how that's done, which I think is important to understand the limitations
of these smartwatches, we can consider this
classic picture of a monitor that you all may have seen on
yourselves, family members, certainly on TV anytime
you're in the emergency room, certainly the
intensive care unit of the hospital
undergoing a procedure, you'll see something like this. To hone in on these two rows, so this bottom row represents a signal from a device called a
pulse oximeter. This is the thing with
usually red light that's placed on
typically a finger, could be placed
on an ear lobe or a toe and the purpose of this is to infer the
oxygenation of the blood. What it's doing is
shining the light and then it has a little
camera and it's looking at how that light is reflected and the way the light
is reflected or the amount of light reflected is influenced by the
amount of oxygen. Then the EKG we
already talked about, that's this top part.
Now, smartwatches
take advantage of the fact that there is a
fairly prominent pulse, right where watch is tend to sit this radial pulses as
shown in this middle picture. On the back of smartwatches
that report heart rates, so Apple Watches,
Samsung devices, Fitbit, they are
using the same light based sensor called
photoplethysmography, referred to in short as PPG. The same technology used
in a pulse oximeter, where it's shining
a light and looking at how that light
is reflected back. Now, although Apple watches now extensively can report
on oxygen saturation, it's actually looking
at the waveform change over time to infer
the heart rate. Generally these smartwatches are again using this
light based sensor on the back that is inferring the pulse from the changes
in the light wave form.
It's not generally, certainly when it's
reporting heart rate it's not using anything electrical. It's not inferring anything
from the actual EKG, just from this pulse waveform. One important and
fairly common way that it can be
fooled is if someone has fairly frequent early
heartbeats such as something called a premature ventricular
contraction or a PVC, that sometimes can
occur early enough that it doesn't generate a
particularly strong pulse. I will see patients for example, with frequent PVCs
that will say, "Yeah, my smartwatch said
my heart rate was 25." When in fact their
heart rate is 50, but every other beat is a PVC.
Similarly, we can infer, or this is the idea, the presence or absence
of atrial fibrillation, given the nature
of the waveform. This is some raw data that we collected as part of
a research study. This is sinus rhythm, this is the
photoplethysmography waveform and normal rhythm and this is
it in atrial fibrillation. In fact, before Apple and Fitbit develop this themselves, we actually published
the first study to demonstrate that, yes, indeed a smartwatch and
we did use Apple watches, this was in collaboration with some data scientists that
subsequently spun out a company called Cardiogram
and they employed a machine learning
algorithm that we trained based on patients with and without
atrial fibrillation.
We tested this in
people undergoing cardioversion
procedures where it proved to be highly accurate, although that was a really
controlled, careful setting. Then we tested it in an
ambulatory population where we noted it still was better
than flipping a coin, but it wasn't terrific at
detecting atrial fibrillation. Then subsequent to
that companies got really interested in
doing this themselves. There have been published
studies now from Apple, from Fitbit and from Samsung. The Samsung study by the way
was done by our group led by our chief for
cardiology, Jeff Orgen.
Now, you might
think, "Okay good, then we're going to detect
more atrial fibrillation. You said that it can
be asymptomatic. You said that it can lead to stroke and blood
thinners prevent stroke, so isn't this all good? We're going to detect people who don't know they have a fib, and that indeed motivated
our original study." Interestingly though,
there's really no consensus that
there should be, at least from professionals and from scientists and
from clinicians, there's no consensus that
we should be conducting population-based screening for atrial fibrillation
among everyone. Now, why is that? This brings us to this concept that we all learn in medical
school that I think is really useful to share and to try to
communicate and frankly, this is all pertinent to
really any evaluation of a medical test and what we call the test
characteristics. This is very classic in any Epidemiology course or
medical student course. We learn about this
two-by-two table where you imagine the disease
here in the columns, presence or absence
of a disease, and you're evaluating
a test whether it's positive or negative.
We tend to talk a lot
about the sensitivity, which is the true positives or A here over the true positives
plus the false negatives, and the specificity
which is the reciprocal, which is the true negatives over the false positives plus
the true negatives. These are not influenced by
how common the disease is, which makes sense because
within this column on the left, they all got the disease, this column on the
right, they all don't have the disease. But the reality is
that when we are in clinical practice and we are confronting a patient
with a positive test, we by definition do not know if they already have
the disease or not. These things, sensitivity
and specificity, they can be used in a
research study where you have some reference board standard that everyone is subjected to.
But in clinical medicine, we have to think about
these other types of characteristics come
from evaluating the rows here rather
than the columns. Again, I'm confronted
with a positive test. I'm interested in the
positive predictive value, which in this case
is A over A plus B. Another way to say that is, given a positive test, what is the likelihood
that test is true? And what's important
related to screening is that these characteristics
in the rows, these predictive values
are highly influenced by disease prevalence and
that's just the reality. I'll give you an example. Let's imagine we have a test
that is highly accurate, so 95 percent sensitive
and specific, which as far as medical Tesco, that's really pretty good. But the prevalence of the
disease is about one percent, and it's not unrealistic
to imagine that in a population of people
wearing smartwatches, which is probably will
skew a little younger and healthier than even
the general population. About one percent
is probably right. Mathematically, given
that prevalence and these outstanding
test characteristics, if you calculate the
positive predictive value, it will be 15 percent.
What does that mean?
That means that 85 percent of those tests
will be false positives. This is one of the reasons that no professional society
has come out to say, "Yeah, we should
screen everyone for atrial fibrillation
because there's going to be a lot of
false positives." Now you might say so what,
why does that matter? Well, that's going to lead to a lot of unnecessary anxiety. Almost certainly quite a bit of unnecessary health care
utilization and even potentially unnecessary and even inappropriate prescription
of anticoagulants, which in general can
provide a lot of benefit to properly
selected people with AFib.
But in healthy people,
there's a risk of, well on everyone there's
a risk of bleeding and if you give it
to enough people, some are going to experience
bleeding where they never really should have received
that drug in the meantime. I was asked to write
a commentary on this Nature Reviews cardiology and pointed this phenomenon out, this new era we're in.
Where if you consider the
conventional way this works, which I went through
on the left, where we first perform scientific research
that undergoes peer review that is then disseminated leads to expert
consensus that informs clinicians and the clinicians inform the patients and
the general public. Now we have this very interesting
situation where we have private industry
marketing devices that are making
diagnoses of a disease, atrial fibrillation
directly to patients. There's no intermediary
here of scientists, clinicians, professionals,
or societies. At the same time, of course, we're trying to do the research
to inform these things. But we are stuck
with this situation. I've argued that we need to do a better job educating
the public, hence, a big part of the motivation
for this very talk and I think understanding what is atrial fibrillation,
why do we care? What do we do about it?
Is really critical and becoming more critical given the presence of these devices.
Now, there's a bit more
to say about them, and that is that they
sometimes are accompanied with the actual ability
to obtain an ECG, such as many Apple watches now, this is also on the left is what's called
a cardio mobile device, which is not a
smartwatch per se, but is a separate
handheld device. They now have a
credit card device that you can hold that
has electrodes that pairs with a mobile app. There are algorithms that
will read these ECGs. They're imperfect, but the
ECGs can be saved and sent to providers and the algorithms are almost certainly accurate if they indicate
a normal rhythm.
Now, this does
raise the issue of potential problems in
a dating providers with a lot of strips. But this may be especially useful for rhythms beyond
atrial fibrillation. The algorithms in general, using the light-based sensor, they will just comment on AFib or something
non-specific or normal. Whereas the ECG, the algorithms won't necessarily call the specific
abnormal heart rhythm. But again, if those PDFs can be saved and shared with a
health care professional, diagnosis can be
made and in fact, this is where I think these
devices are pretty clearly useful and that's in patients with an established diagnosis.
If you have a population
of people who already have received a diagnosis
of atrial fibrillation, then your prevalence
is very high, so false positives for the same mathematical reasons I just described are much lower. This can help people
with atrial fibrillation decide when to take what's
called the pill in the pocket. Some patients are prescribed a medicine to take to
convert their AFib, but the symptoms
aren't always clear. Someone may just feel a little off or a little anxious
and they're not sure. Is this AFib or am I
just feeling this way? Is it worth taking
one of these pills? These devices may be helpful
there or to determine if a particular drug
they're taking on a daily basis is working, whether they should have another catheter
ablation procedure, or to identify various triggers. Oh, it's when I consume alcohol. Yeah. The device really is
showing atrial fibrillation. Moving on now to various treatments for
atrial fibrillation. There's a big dichotomy here. These things aren't necessarily
mutually exclusive, but we often discuss
them that way, and that is what we would
refer to as rate control, meaning allow that
person to remain in atrial fibrillation
versus rhythm control, meaning try to get them out of AFib and keep
them out of AFib.
Rate control the goal
is really to work on that AV node to
make sure it's not allowing the AFib to make
the ventricles go too fast. Now there were several
randomized trials that were done now
about 20 years ago that counter-intuitively
perhaps failed to demonstrate a clear benefit of the rhythm control strategy, meaning that let's get rid
of the AFib and try to suppress it over rate control. However, there are
several limitations to those studies that
are worth emphasizing. First, there wasn't really any evidence that the people who underwent the rhythm control generally did worse so they were considered
fairly equivalent.
Second, the means to maintain sinus rhythm
were sub-optimal. We've generally
recognized that this was before catheter ablation and
many of these people were not necessarily treated by electrophysiologist
that understand the various nuances
of various drugs available to suppress a fib. It's also very likely that the most symptomatic
patients were not enrolled. If you're a treating
physician and you have a patient that you
know feels horrible, and atrial fibrillation feels
great and sinus rhythm. You're not going to be very likely to encourage them to enroll in this
randomized study.
Most of the studies
were quite small, and then the duration
of follow-up may have been insufficient to really see the adverse
consequences over time. Now a really
important lesson from these trials that relates back to my initial discussion of the formation of blood clots in atrial fibrillation is that a normal sinus
rhythm strategy is clearly not sufficient
for stroke prevention. How do we know that?
Because in these trials, especially the main one
that's called a firm. Per the protocol, if someone was
randomly assigned to rhythm control and they came
back to clinic and yes, they were normal rhythm, the practitioner could stop their blood thinner
and it turns out that their stroke rate was the same as the people in
atrial fibrillation. Now you could argue that well, but it was still lower
because they were authentic coagulation
and indeed, there has been a question, well, if you have normal rhythm
plus anticoagulation, would that really knocked
down the rate of strokes and other thromboemboli and indeed
that ends up being true.
Much more recently, this
trial was published, really the first in
many years to revisit this question of
rhythm control versus rate control in the
modern era with ablation, with electrophysiologist
caring for these patients and
really importantly, with the idea that
anticoagulation or blood thinning
would be continued even in the people randomly
assigned to a normal rhythm and they found that
really in every outcome, there was less problems in the people randomly assigned
to the rhythm control, meaning let's try to
prevent AFib arm. The main outcome that was statistically significant
was a composite, but really each of
these is lower, so there was less death that was statistically
significant, less stroke interestingly, despite the fact that
everyone got blood thinners, less hospitalization, either for heart failure or
essentially a heart attack. But sometimes a rate
control strategy is still reasonable depending on various factors
and we are slowly moving towards a rhythm control
as more of a first line. One of the important caveats of that study I
just described is that these were in people with relatively new
atrial fibrillation.
The one thing that everyone has in our field of
electrophysiology has agreed is it makes
sense to pursue a normal rhythm control
strategy in people who have symptoms in setting
of atrial fibrillation. If we imagine we've decided we're going
to do rate control, we're not going to
try to suppress the atrial fibrillation or we can't suppress the
atrial fibrillation, which does sometimes happen. There are a couple of drugs
to help slow the AV node, and those include beta-blockers.
These are these medicines
that end in olol like metoprolol, atenolol, etc. Calcium channel blockers
can do the same, specifically diltiazem
and verapamil, not all calcium channel
blockers per se. Both of these classes of drugs
do decrease blood pressure so that can be a win-win when a patient already
has high blood pressure, but can be limiting or problematic in people whose
blood pressure is too low. Digoxin is also useful here, perhaps is considered
more second line because it
doesn't work as well.
Also it may actually promote atrial fibrillation in those who have intermittent episodes, which is less of an
issue in people who are in persistent atrial
fibrillation. Now, all of these drugs, they've worked on the
AV node to slow it. That's the intention. When someone's in normal rhythm, they will also slow
the sinus node. This can be a problem when someone's in and out of
atrial fibrillation, they're going too fast in atrial fibrillation hence we give them these drugs
to slow them down, then they convert to
normal rhythm and now they're going to slow
in a normal rhythm. Usually that's not
dangerous per se. but people can feel really
tired because they can't mount an adequate heart rate
response to exercise. Sometimes people can pass
out when they convert from atrial fibrillation
and it takes a while for that sinus node to recover. This is something called Tachy-Brady syndrome
or too fast, too slow syndrome and that is a indication for a
pacemaker which works very well to address
this slow rate. I want to talk about the
pacemaker because of that and also another solution for many
with atrial fibrillation.
Putting in a pacemaker
is very straightforward. It does not entail having
to open the chest. We make a very small
incision in the upper chest, just down to the muscle layer, just under the skin, we make a little pocket. We go into a vein here, much the same way that we go into a vein
in the legs to do a catheter ablation or an operator might go into an
artery to do an angiogram. The way we do that
is usually now under ultrasound or some
other visualization. We'll place a needle
into the vein, get a little blood back, put a wire through that needle, take the needle out and
put a little plastic tube, essentially like a large
IV over that wire, take the wire out
and then we can introduce either catheters
or in this case, pacemaker leads that will
look like this through those veins down into the
heart under x-ray guidance. There are a number of
ways to then fixate those leads to the heart via, for example, a screw that we can deploy with a wrench on
the outside of the lead.
We connect the leads
to this generator, tuck it in that pocket and
sew it up but essentially this is a very common,
straightforward procedure. Now, in general, all the pacemaker can do is prevent the heart
from going too slow. It just paces the heart, activates the heart, preventing
it from going too slow. It doesn't really otherwise, slowdown rhythms that are too fast, like atrial fibrillation. There's one exception and that is when patients
have something called atrial flutter, which is related to
atrial fibrillation but is a more organized
single circuit. There is one type of
pacemaker were in the lead in the upper
chambers so this lead, by the way, is in
the right atrium. This lead is in the
right ventricle. It can sense that fast
rhythm pace a little faster and sometimes help break that circuit in the
setting of atrial flutter. Now this is pertinent to this
talk on atrial fibrillation partly because of that
Tachy-Brady syndrome I described, but also because of this very
straightforward procedure called an AV node or
AV junction ablation. This is when we have a
patient that we're trying to rate control and we can't because the
medicines aren't working.
Sometimes we pile
on these medicines, their pulse is still too fast or their blood pressure
is just too low on those medicines or
sometimes these are patients where
we've really tried to suppress the
atrial fibrillation, but the medicines don't work. We try catheter
ablation, doesn't work or things aren't tolerated, then we place a pacemaker, as I just described. We do this actually very
simple procedure where we go in with a catheter
and burn the AV node, which essentially electrically disconnects the top chambers, the atria from the ventricles. We allow the patient to be in atrial fibrillation
but the ventricles, the pulse, they just
don't know it and they never go too fast and
they don't beat irregularly. They are, however, then
dependent on the pacemaker. Now I should mention the
first ablation ever done in a human for an arrhythmia was done by a doctor Melshaman, a colleague of Dr. Dolan's and myself and really
a mentor of ours. It was done at UCSF in 1982
and it was this procedure.
They used a different
form of energy, but it was essentially this AV junction or
AV node ablation. Moving on to other
therapies and imagining now we do want to obtain and
maintain a normal rhythm. A common way to do this in those with persistent
atrial fibrillation. It's not breaking, they're
not reverting on their own to normal rhythm is to do what's called
a cardioversion. This is the same
procedure you see on TV where they put
on pads and they yell clear and they
deliver a shock that is usually in a scenario of
a ventricular arrhythmia, a deadly arrhythmia,
in this case, is a very calm. We really do help clear after charging and before we shock but it's a very
common procedure. We often do four of these
on any given day at UCSF and hospitals all
over do this frequently. It works very well to convert someone from
atrial fibrillation to a normal rhythm
but it doesn't do anything to prevent the atrial fibrillation
from coming back.
The most important determinant
of the success tends to be how long the atrial
fibrillation has been going on. If it's been present
for years and years, sometimes even the
cardioversion doesn't work. We can sometimes facilitate
that by adding medication, sometimes trying an ablation but it usually works
and the challenge tends to be the recurrence of the atrial fibrillation
down the line. That brings us to various
drugs called anti-arrhythmic drugs that are designed specifically to suppress
the atrial fibrillation. There are several. They all have their
pros and cons. None of them work perfectly, of course, for any
given individual, they may work
especially well but in a general population with
atrial fibrillation, you give these drugs, 50-70 percent of
them will be able to maintain a normal
rhythm at about a year.
Certain drugs are contra, indicated in the setting
of certain disease states, especially in people who have any blockages in the arteries
supplying their heart or a history of heart attack or heart failure then
we're a little bit more limited in terms of the
drugs that we can give. Now, the other option as I've referred to is
catheter ablation. This is a more complex ablation than simply burning the AV node. The goal here is to render one free of atrial fibrillation. The way this works is we do something called a
transseptal puncture. This comes back to my
disclosures I mentioned of that a company Baylis they make tools to help us traverse
the septum and the point of that is to burn or
freeze the opening, the ostia, or just outside these veins that
come from the lungs, hence pulmonary, into
the left upper chamber. I should mention,
these veins have little sleeves of muscle
that conduct electricity. If we put a catheter like this, what we call lasso catheter that has these electrodes on it, into the pulmonary vein we
can measure signals that are conducted from the
rest of the atrium.
It's been demonstrated that if one can electrically
isolate those veins, generally by burning around them or delivering this what's
called a cryo balloon, essentially freezing the
tissue surrounding them, you can free the majority, not everyone but the majority of people with
atrial fibrillation. Now, how does that work? What are these pulmonary
veins have to do with a fib? That's actually an
area of ongoing study, but there are several
potential explanations and they may vary depending
on the individual person.
This approach was really first described by clinicians
in Bordeaux, France that noticed that some people with interim-in-nature
fibrillation seemed to have the rhythm triggered
by these early beats called premature atrial
contractions originating inside the pulmonary veins. The first procedures they were going in for the pulmonary vein, they were trying to hunt down
these rare early beats and burn them and they showed
that that could work. But the problem was then other premature atrial
contractions which show up elsewhere in other
pulmonary veins and burning inside the vein could lead to
narrowing of the vein, which led to problems with
shortness of breath, etc. That subsequently
led to this idea of, maybe we just need to
electrically isolate all the veins and none of
those triggers can get out. That's one potential
explanation. Another there is what's called a Cox-Maze procedure
developed by a surgeon with the theory that atrial
fibrillation represents these partially
re-entrant wavelets that require a certain
amount of tissue to perpetuate and to be sustained such that if
you cut up the tissue, cut up the atria,
you created lines of scar or electrical block these
wavelets couldn't coexist, and therefore normal
rhythm would take over.
It's possible that
with our ablation, we're also making the
atria a bit smaller, providing less room
for these wavelets. Then third, there's this idea that maybe we're influencing the autonomic tone that I referred to earlier
specifically, there are nerve fiber
bunches where it's mainly the vagal
anti-adrenalin input on the heart that happens to sit right next to and right outside where we tend to burn or freeze when we're
trying to get rid of atrial fibrillation and
isolate the pulmonary veins.
Indeed sometimes with these
procedures as we're burning, we'll see the heart slow
way down because we're irritating that vagal bundle
of nerves called a ganglia. It's not uncommon after an
eighth of ablation that individual's heart
rate to go up a bit and we think again, that's because we've disrupted this vagal atrial connection and there's some evidence also that if you
activate those ganglia, especially in animal models, you can induce
atrial fibrillation.
It may be that one of
these things is active in various patients or
it's a combination. Again, the point is that we know empirically in randomized trials that this does tend to work. This ablation procedure is the most effective means
to maintain normal rhythm. In studies that have
compared this to drugs, it tends to win out.
Interestingly, it's
especially useful in heart failure patients
with some evidence, especially those
whose hearts are weak rather than simply stiff. With some evidence from at least one
randomized trial that ablation in those patients can
actually reduce mortality. The success rate is
probably on average 60, maybe 80 percent, perhaps as high as 90 percent. Repeat procedures probably do
increase that success rate. The risks are generally low
altogether 3-5 percent. But importantly, the
great majority of that 3-5 percent are things that either take
care of themselves, they're not going to result
in permanent disability. The risk of death,
permanent disability, something that's really
going to be negatively life-altering is substantially
less than one percent. I have become much more
sanguine about this procedure. The last time I spoke
about this in 2015. I'm certainly doing
this procedure. I've been doing it
since I was a fellow, I started my fellowship
18 years ago. I've been doing it ever since I started on the
faculty 16 years ago. I have been performing
this but I've tended to err on the side of
being quite conservative.
I tend to approach things with some healthy
skepticism and we'll say that the tools
have really improved. They've made the procedure, I think safer, more effective, faster and I've seen more and more anecdotal
success that has accompanied randomized trial
data demonstrating good results with this
in select patients. It's not necessarily indicated
for everyone and again, happy to talk more
about that in the Q&A. The best candidates
are those that are symptomatic and especially those that are symptomatic and
that have failed drugs. The individuals who are not the optimal
candidates tend to be those with a very long-standing atrial fibrillation
for years and years. Sometimes it can help
those individuals, but we know the success
rate is not as good. Older patients tend to
experience more risk factors. Once we get into the mid, certainly the 80s, and definitely, mid 80s, 90s, I would tend to shy away
more from this procedure.
Now, very important to talk about preventing
thromboembolism. This means, by the way, it's not just stroke, which is what we mainly talk about with
atrial fibrillation appropriately because it's
so important but remember, there's no reason
these blood clots necessarily just
go to the brain, there's now evidence
that people with a fib are at a high risk
for heart attack worsening kidney disease. It may also not just be these big clots that
lead to a stroke, but perhaps small
or tiny clots that aren't manifest in any way
in an immediate sense, but over a long period of
time may lead to dementia. Really this can lead to harm to various organs and prevention of thromboembolism is meant to prevent all of these things.
The mainstay there
is to use what we call anticoagulants or a particular type
of blood thinner, which are highly effective in preventing these
complications. Randomized trials have
shown less death with them, as well as substantially
less stroke and these other complications. Now we use the term
anticoagulant, mainly in distinction
to what are called anti-platelet drugs
such as aspirin or Plavix. Those are also blood thinners and that they block platelets. This blood clots can have within them platelets
and clotting factors, certain types of blood clots, especially those
for example that form in coronary arteries from a ruptured
cholesterol plaque that lead to a heart attack. Those tend to be quite
platelet rich and therefore drugs like aspirin or Plavix are useful for that. The sort of clot that forms in the setting of
atrial fibrillation, seems to be much less
platelet related and more related to these
clotting factors that these anticoagulants block. Therefore, aspirin and Plavix don't work particularly well to prevent all
these complications and atrial fibrillation. Interestingly, with these
new anticoagulants, the risk of bleeding however, is about the same as with
aspirin, for example. The first-line drugs
we're talking about here, which are also
referred to as no acts for novel oral anticoagulants.
NOACs as opposed to the older, mainstay of warfarin
or DOACs for direct oral anticoagulants
include Apixiban, Dabigatran, Rivaroxaban,
and Edoxaban I've listed the trade names there
as well as Eliquis, Pradaxa, Xarelto, Sayvasa. Warfarin is also an option. Warfarin is also
called Coumadin. It does work well, has been studied in
large randomized trials, but isn't considered
second line to the drugs listed above
due to inconvenience since you have to
constantly check for how thin the blood is and potentially
adjust the dose. There's a higher risk of
bleeding inside the head with Warfarin or Coumadin compared
to these newer drugs. There are some circumstances where the newer drugs, again, those listed above are not indicated in the setting
of atrial fibrillation, such as when someone has
what's called mitral stenosis, which is almost exclusively caused by rheumatic
heart disease. That's a narrowing of
the mitral valve or mechanical heart valves or sometimes people
do develop clots while on these perhaps
more gentle NOACs or DOACs and then we give
warfarin and can then dial up how thin the blood is.
I just wanted to address some common concerns that
seemed to come up fairly frequently and that I hear from patients especially when I first introduced the recommendation
to start one of these. One is, while they're new they haven't
been studied enough. My response is they've been
around now for a decade. In the randomized trials alone included 10s of
thousands of people. Now we have data from literally millions of
people in the real-world, all with very
consistent findings of the benefits of those drugs.
Another concern as well,
they're not reversible, but there are now
antidotes available. Frankly, even before
the antidotes we could control the bleeding. These are gentle enough
that we are now performing procedures on uninterrupted,
no NOAC or DOAC. Oftentimes when we
put in a pacemaker, certainly pretty much every time we do an
eighth of ablation, including puncturing the veins, puncturing the septum between the top two chambers were
doing it in people who have taken their eloquence
or Xarelto that morning. In general, when indicated, the risk of not taking them exceeds the risk of taking them. There's a saying that I learned from older investigator
that I really like, which is patients don't
call you in the middle of the night to thank you
for not having a stroke.
Sometimes it's too easy to imagine why I'm doing this
thing it's going to lead to bleeding and it's
harder to fully appreciate the absence
of something happening. But of course, we can
all recognize that's not necessarily rational per say just because it's
easier to imagine doesn't actually mean it's more probable and the data clearly show that generally these tend to lead to better outcomes. Now I mentioned properly
selected people. What do I mean by that? Even among people with
atrial fibrillation, there's clearly a spectrum
of risk of stroke and thromboembolism and we use the scoring system
called the CHADS VASc. We used to use
something called the CHADS2 score but in general, younger patients and those without any other
cardiovascular risk factors. No high blood
pressure, no diabetes, no history of stroke, no heart failure maybe at sufficiently low
risks that it makes sense to not prescribe
a blood thinner.
Probably especially
selling those who have fairly rare
atrial fibrillation. I will say that given
the convenience of these newer drugs and
how well they work, the field seems to be moving
a little bit more towards erring on the side of
generally recommending them. In the next few slides, I'll move to this new way to approach atrial
fibrillation that has to do with prevention, which of course is
always preferable rather than delivering
some therapy for a disease that someone
has to first suffer not to mention the hassle and the risks of the
various therapies.
I will say it's been
very interesting when I was a medical student
and even a resident, we didn't think about
preventing atrial fibrillation, we didn't think about preventing electrical problems we certainly thought of that when it
came to heart attacks, reducing cholesterol,
stopping smoking, but atrial fibrillation was just bad luck could
just happen to people. More and more, we're
starting to recognize that perhaps at least in many, probably not everyone, especially those for
example who have strong genes that make
them susceptible. But in many, we might be able to prevent the disease and also use lifestyle factors to reduce
the risk of recurrence. In fact, these are figures from on the left the
American Heart Association, which now considers Lifestyle
Risk Factor Management one of the pillars of treating atrial fibrillation along
with Anti-coagulation, Rate Control, Rhythm Control. On the right here,
this is a statement regarding research priorities
from the NIH, the NHLBI.
I was fortunate to be
part of this group related to secondary prevention
of atrial fibrillation. Meaning that once it occurs, what are the things we can
do to actually prevent it from happening with a
major focus on lifestyle. One of the leaders in this
has been a colleague, Sanders, who's in Adelaide, Australia, where they do a lot of oblations and they
had a long waiting lists. The story goes that they add
all these patients that were waiting a really long time for their ablation and they
wanted to help them, wanted to figure out what
can we do in the meantime. They ended up examining the effect of fitness
training and weight loss, especially in obese patients.
This is a very famous
randomized trial where individuals are
randomly assigned to a Weight Reduction and Cardiometabolic Risk
Factor Management Strategy versus just letting
them be and they showed that indeed if
they can lose weight, become more fit, that their atrial fibrillation
burden went down substantially. Now alcohol has long been considered a possible trigger
for atrial fibrillation. This was a meta analysis
now over a decade ago, and you can see that the findings and
individual studies with equivocal will be where
this solid line of one is favoring a relationship
between alcohol and AFib. and we'd be on the right
side of this line. One of these studies
was mined from actually some work I did
as a fellow at UCSF, but the general conclusion
was those who tend to drink more tend to be at a higher risk of atrial fibrillation which
then leads to the question, well, if you've been
drinking is the cat out of the bag or can you still
influence the risk-free of it? We published this study
where we leveraged a large prospective
NIH funded cohort and we looked at the people who used to drink and then quit, and found that those
people actually did have a lower risk of afib compared to those that
continue to drink.
This was one of several studies that
the first author here; Aleks Voskoboinik, who
we had the privilege to help train in his cardiac electrophysiology
training is Australian. A very clever study that he did and he's
kindly referenced the slide I just showed is one of the inspirations
for this study. So this was part of his PhD with another
electrophysiologist named Peter Kistler in Australia where they took patients with atrial fibrillation who
drank pretty heavily, and they randomly assign them
to try to convince them, try to stop drinking versus didn't really
tell them anything. They continued to
drink. Those randomly assigned to abstinence exhibited a substantially reduced risk of recurrence and atrial
fibrillation burden.
Now what are the
mechanisms here? What's going on with alcohol? There's this common conception that alcohol is
good for the heart. There is some
evidence that alcohol when consumed in moderation; meaning about a drink a day, there may be beneficial
cardiovascular effects. There could still be a
reduced risk of AFib. This is very controversial,
really unknown. I'm happy to talk
about this in the Q&A. We're trying to get a
major study funded here when not consumed in excess, but there's also evidence
that in people who already have afib that even a
drink might be enough. I'll tell you about that
a little bit later. Now the protective effects
tend to be attributed to maybe beneficial
effects on lipids, and maybe it thins the blood a little bit which
helps prevent heart attack, but the mechanism by
which alcohol may lead to atrial fibrillation has
largely been not understood.
We did a randomized
study where we took patients undergoing
afib ablation. They had catheters in
there hearts so we can measure the electrical
activity and characterize the electron activity inside the atrium and we randomly assign them to a [inaudible] of alcohol titrated to a breath
alcohol concentration of 0.08 percent versus a
double-blind must placebo. We found that in
the alcohol group the refractory period of the pulmonary vein electrical
activity became substantially shorter only when people were exposed to alcohol not
in the placebo group. That shortening of atrial
recovery is something that is thought to render the atrium
more prone to fibrillate.
If you imagine those
multiple wavelets I described more of them can co-exist if the tissue is
recovering more rapidly. We thought it was
especially interesting that this was found on
the pulmonary veins which are the target
of our ablation. We also showed in
ambulatory people with paroxysmal atrial fibrillation
that if we put sensors on them and continuously
to monitor them for their heart rhythm as well as
an alcohol sensor that when they consume the
alcohol their risk of a discrete atrial
fibrillation event went up substantially
in the next few hours. This was really the first time that these acute
effects were shown. All the previous data had to
do with long-term effects, alcohol as a risk factor for the development of the disease or the
burden of the disease but this was the first
evidence that there may be immediate effects
of alcohol on AFib. The hopeful thing is if that's true that also suggests
that if you stopped drinking alcohol then fairly
immediately you're going to reduce your risk for
a given AFib episode. It was very gratifying
to see that study I just described was covered in
The New York Times and they talked about very
generally alcohol affecting the heart
even though it was about atrial
fibrillation whereas in the past I think
again all the alcohol, heart disease media
attention has focused on heart attacks as well as strategies to
prevent the disease.
Just to cover a couple of
other lifestyle factors. Smoke along with many
other horrible things that can happen from
smoking tobacco, there's now evidence
smoking also does increase the risk
for atrial fibrillation. We showed that actually secondhand smoke can be a risk factor for
the development of atrial fibrillation and
then we just recently published this study
looking at records from more than 23 million
Californians seeking healthcare and found that various substances:
methamphetamine, cocaine, opioids,
and even cannabis, actually increase the risk
for atrial fibrillation. You may be lamenting, Oh man, I can't drink alcohol, I can't use cannabis, or Dr. Marcus is telling me
I shouldn't use cocaine, that's all fairly pretty true. What about my coffee? Don't take away my coffee. I do have better news here. There is this
conventional wisdom that coffee leads to a
arrhythmias and there's even professional society
guidance without really substantive
data that warn against caffeine consumption
to avoid arrhythmias. We were interested in
looking at this and just looked at common
early heartbeats. This is in a population-based
NIH funded study.
We could find no
relationship between how much coffee one
said they tended to consume and the frequency
of those early beats, and then more
recently we published a study using data
from the UK Biobank. This is from several hundred
thousand participants in the UK. Actually it found
that those who drink more coffee experienced
a lower risk of various arrhythmias, and that lower risk
was largely driven by a lower risk of
atrial fibrillation. Now these are observational
data and not randomized data. We tried to address that by performing something called a Mendelian randomization studies, so we looked at caffeine
related genetic variants that are at least ostensibly
somewhat randomly assigned. There we didn't find a heightened risk of
atrial fibrillation, but we also didn't
find a lower risk using these genetic markers as surrogates for
random assignment to at least
caffeinated products.
I will end here with these conclusions and
then look forward to hearing what you're curious about and hopefully
answering your questions. In conclusion, atrial
fibrillation is very common. We make the diagnosis
using a electrocardiogram, also called an ECG or EKG. Not to be confused by the
way with a echocardiogram, and a echocardiogram that's the ultrasound of the heart and that electrocardiogram
needs to be over-read by healthcare
professionals. We shouldn't rely on smartwatches by themselves
to make that diagnosis. Smartwatches can identify those
with atrial fibrillation, but we still need to
figure out how to optimally use them in
the general public. They perhaps have the
greatest utility among those who already have a
diagnosis of the disease. The goals in treating atrial fibrillation are to
improve quality of life, to avoid adverse remodeling, weakening the enlargement
of the heart, and to prevent stroke
and other thromboemboli.
There are many effective options to address each of these, often working in concert. Now we recognize that
staying physically fit, especially among
those who are obese, as well as minimizing alcohol
and avoiding tobacco and illicit drugs may all reduce the risk of
atrial fibrillation. When it comes to coffee, I'm not recommending
anyone start to consume coffee because there's no randomized control
trial evidence for that. However, if you
enjoy your coffee, then I certainly would not
discourage its consumption.
That is all I have. Thank you
so much for your attention. I hope this has been
interesting and useful. Great talk, Dr.
Marcus. Thank you. We go through some of the
questions here in the Q&A. There's a number of questions regarding the risks to some
of these procedures that you described with regard to atrial fibrillation ablation here with device implantation. Maybe you can just touch on what are some of the risks of atrial fibrillation ablation? What are some of
the risks of poking a hole to perform
this procedure? Maybe two holes in the heart. Any problems with these leads in the ventricle of the heart in traversing a valve when we put in a pacemaker
defibrillator? Great question.
The most
common complication of the ablation procedure is
where we go into the veins. That can result in
various issues. Again, that makes up maybe 2-3 percent of all of
those complications, perhaps even as high
as four percent. That rate, by the way, is probably now lower. We used to do this based purely on what we
knew about anatomy.
We would feel the pulse of the artery and we would go
next to it to get the vein. Now we use ultrasound, so we're visualizing
the vessel as we go in. We use very small needles what
are called micro puncture needles and very small wires to initially
access the vein. Usually, if that's a problem, it will manifest
as a collection of blood or even a connection
between an artery and vein.
Many times those things just
take care of themselves. If that becomes an
issue, usually, the scenario is not generally
the patient is unstable, but rather they go home. They say, "Yeah, the pain there is not getting
better,'' or are they notice it's swollen. We get an ultrasound, we consult our vascular
surgery colleagues. They will often say just keep an eye on it and get another
ultrasound in a few weeks. Once in a while, about less than one percent
of the time they'll actually have to
go in and fix it. In terms of other complications, the poking of the hole
between the septum, between the right
and left atrium that we do deliberately, that seals up in about, we think it takes about a month. In fact, we published
a paper a long time ago showing that when you
do a repeat ablation, getting back across a little
bit more difficult so that it heals up
really very well.
There is a very
interesting phenomenon that hasn't been
very well studied, but it's been pretty extensively described that in
that first few weeks, some people can
develop migraines, especially visual migraines, and there's also evidence
that people with migraines may more often
have a little hole there. We all have little holes
there when we're in the womb, and then when we're born, there's a little
flap that closes. About 20-30 percent
of the population, there's still a
little hole there.
We don't necessarily go
through that exact spot, but very close to it when we
do our transseptal puncture. The idea of these migraines, that mechanism that's purported, is there some substance that normally is filtered by the lungs that comes back to the right
side of the heart, when there's a
little hole there, it goes directly
to the left side and that causes the migraine. Interestingly, those symptoms reliably go away after
three or four weeks, demonstrating that we think
that that hole is closing. Now one of the other more
feared complications that remains relatively common, maybe as high as one percent, is a unintentional
hole in the heart, which usually manifests as
bleeding outside the heart, which patients undergoing these procedures
are at high-risk for because we're giving
very potent blood thinning, because we're on the left
side of the heart and we want to avoid any blood
clots forming there.
Should that occur, the solution that works
the great majority of the time is we enter in just under the rib-cage
to drain that blood. That hole usually just seals up again and the key is just
draining the blood from outside the heart or getting into the sac
that lines the heart. Usually those patients, they spend the night
with a catheter sitting in there and that
can be uncomfortable. But the great
majority of the time, that's less of an issue.
Another final complication
I'll mention that is really the worst complication and is extraordinarily rare and it's so rare that it's actually
quite hard to study. To my knowledge we've not seen this among any of our
patients ablated at UCSF, although we certainly care for these patients at UCSF that are transferred from elsewhere, where a hole forms between the esophagus,
the swallowing tube, which, as I described,
is directly behind the heart and the back
of the left atrium.
One of the things that is especially pernicious
about this, or it makes it really difficult, is that that doesn't
form immediately, it comes up three weeks later and can present in unusual ways. And so we take several measures to make sure we really try
to avoid that. These days, and different colleagues do
some different things. I've been using a device where we put down actually
a device that cools the esophagus and prevents
it from getting heated up while we're burning
the back of the heart. To quickly move on to the
pacemaker question again, I will defer most of
that to Dr. Dolan's talk in hopes of motivating
many of you to attend that, but most of the time
it's not an issue. The heart is unbelievably forgiving of many of
the things we do, but it can sometimes lead to difficulty about
tricuspid valve that we traverse with the lead that goes into the lower
chamber closing and that can be an issue sometimes usually over
a long period of time.
How about cardioversion? There's question
on that as well. Yes. The risks of
cardioversion are very low. It is an incredibly
safe procedure. It's been around for
a very long time. The most common adverse
consequence of it are skin burns, which again are mitigated by the nature of the pads we use. We see that very rarely. The other risks would be shocking the heart into
another dangerous rhythm, and that can broadly
take two forms. One is we can shock into a
dangerously fast rhythm. There is something that we do to synchronize the shock with the timing of the way that the heart depolarizes
to prevent that. But in that worst-case scenario, the solution to
that is essentially to shock the heart again. There is also a risk of shock the heart of someone
in atrial fibrillation and their normal sinus rhythm
comes back very slowly, and then we would give medicines
to speed up the heart.
We can also use those
same pads actually to pace the heart from
outside if needed. A couple of questions regarding stroke and
atrial fibrillation, including any concept of
how or any idea of how long it takes for one of these clots to
form in the heart? Then with catheter ablation, do we have any idea of whether the ablation procedure long term helps impact the risk
of stroke or lowers the risk of stroke with those individuals that
have an ablation? Yeah, two great questions.
I'll answer the second
one first I think. The honest answer
is we don't know. There's some controversy
in the field. There is observational data, so there's not yet randomized
controlled trial data which we try to rely on
the randomized data, mainly because that's
really the only way to fully mitigate
against confounders. But there's some observational
data that those who undergo ablation may experience
a lower risk for stroke. Some will sometimes then
stop anticoagulants.
In my practice, if someone's stroke risk based on their other factors
is sufficiently high, even if all the
evidence is they have just a normal rhythm I tend to favor continuing
a blood thinner. If their stroke risk
isn't terribly high, I'll have a discussion with them sometimes obtain a monitor
to make sure there's no asymptomatic afib and what we call shared
decision-making with the patient, I will sometimes stop
the blood thinner. I usually give it about a year though to make sure it's gone. Then there are the people
who, as I mentioned, are so low risk of
stroke that it's not clear that they would even need a blood thinner
and those folks, I do often stop it.
Sorry. What was the
first question? How long does it take
to fully recline presumably once an individual
is in atrial fibrillation? Thank you. There is a convention that we think it takes about 48
hours at a minimum. However, there are
cases that have been described where the evidence is that atrial fibrillation
has been present for a shorter period of time
and yet strokes can occur. Interestingly,
cardioverting the heart heightens the risk for stroke, mainly because
after cardioversion the moving from atrial
fibrillation to rhythm, that seems to stun the left atrial appendage and make the blood even
more stagnant. That is not a property of
the electricity we deliver. It's a property more
of the conversion from the atrial fibrillation
to the normal rhythms. Whenever we do a cardioversion, especially if it's been
more than two days, we always make sure
someone has to be on uninterrupted
anticoagulation for at least ideally a
month afterwards. Now, it's also
important to mention, and I tell my patients this frequently that we don't
think just as I described, that it may not
always be the afib that's directly
causal to the clot.
It's also not the
case if you take someone for example that has paroxysmal atrial
fibrillation that, now they're in afib, "Oh no, now I'm going
to have a stroke." It doesn't work like
that and I don't want people to be alarmed when they're in
atrial fibrillation. That's suddenly now at that time they're at this very
high risk for stroke, I think it's more helpful
and more reflective of the reality to think
of it more like treating high blood pressure
or treating diabetes, that the reason to take a blood thinner in the setting
of atrial fibrillation, even when it's coming and going, is to reduce the risk of
stroke over years and years.
Everyone with
atrial fibrillation and they're not on
anticoagulation, on average their stroke risk is about five percent in a year. It's not five percent for a given afib episode is
certainly not 20, 30 percent. We're talking a
longer timeframe. It's important to understand,
as I mentioned earlier, these are not emergencies when
you have an afib episode, you don't need to rush
to worry about it. It's treating it over
the long period of time. A question came up about alcohol and I think
this is on a lot of people's minds and you've done a lot of research
in this area. There's pretty good evidence that heavy alcohol consumption increases the risk of
atrial fibrillation. How do we weigh mild alcohol consumption
given that there is perhaps some data to suggest that
it may be reduces things like heart attacks or
some forms of cancer.
How do we weigh the overall
risks and benefits of it when atrial fibrillation is the problem but there's
all these other things that we have
to worry about? Fantastic question. We don't know and it's, I think very important. There's even some evidence, as I may have alluded to. There's one big
observational study that suggested that in
people without afib, those who drink just one drink
a day have a higher risk of afib compared to
those that don't drink. That was from
continental Europe. I think Germany, but I
don't remember the country. There's other data from
the UK that people who drink one drink a day may have a lower risk of
atrial fibrillation.
As Dr. Dolan mentioned, there's evidence that people
who drink a drink a day, there's even a randomized
trial evidence may have a lower
risk of diabetes, and they may have a lower
risk of leukemia lymphoma. They may have a higher
risk of breast cancer, colon cancer if there's a
lot of uncertainty there. We've actually, I mentioned
this, I think earlier. We're trying to get a
major study funded that would actually compare a drink
a day to near abstinence. A colleague of mine at Harvard Ken Mukamal who is also really a leader in
alcohol-related research, and I have to keep trying for this one grant mechanism
that would work for this.
Unfortunately, and we don't
want industry funding, we want to be
completely unbiased. We're trying. It's
a great question. We think it's
incredibly important and we're trying
to do that study. We would love to do that study
and we think it could be super informative and helpful to countless numbers of people. I would mention more
than two drinks a day, almost certainly harmful even when it comes to
heart attack and we've shown that when it
comes to all these outcomes.
I tend to recommend
to my patients without atrial fibrillation.
I tell them, I don't know. I would generally recommend
not consuming more than a drink in a 24-hour
period on average. Have you noticed that some of these medications
over time become less efficacious and what happens to the patient whose
maybe failed medications, failed a catheter
ablation procedure. Is there anything wrong? Yeah. Treating
atrial fibrillation is often very iterative. It's really a trial
and error thing. I really try to
also emphasize to my patients that their
quality of life is so important because I know it can often happen that we
prescribe a medicine, patients don't feel
well on that medicine, but they say, well, my doctor said I got to take
this medicine.
I always encourage them,
please let me know. If you're not feeling well, the whole point of this is
to help you feel better. Now, the anticoagulants
they don't necessarily help you feel better say hopefully
preventing a stroke. But day-to-day, that's
not their purpose. The day-to-day of the
drug is to suppress afib, they really should be
helping you feel better. If they don't, you need to
try something else and yes, it's interesting
for some reason.
First of all, some people's
given atrial fibrillation just never is susceptible
to a given drug. Sometimes a given drug
will work for years and years and then the afib
comes back despite that. Then we go, it's iterative. We can then try a
different drug. We can then try ablation. Sometimes you try ablation. Patient does really well
for a couple of years and the afib comes back,
then you try drugs. As I say, it's a fairly
constant trial and error, I say fairly because
there are some patients who are on the same
drug for decades, never have any more afib.
Or we do a catheter ablation, no drugs and they're free
of afib for many years. [MUSIC].
