Smith – Management of Refractory Distributive Shock

Dr. Lane Smith is a critical care medicine specialist at Atrium Health Carolinas Medical Center located in Charlotte, North Carolina and serving as the Charlotte academic branch of Wake Forest […]

Episode Transcript

(00:00):
My name is Lane Smith. I'm
a critical care medicine specialist,
down at Atrium, Carolina's Medical Center in Charlotte,
North Carolina,
and we're kind of the academic hub in
Charlotte of Wake Forest School of Medicine.
I work mostly in the medical cardiac ICU,
a little bit of neuro. I train largely
in cardiovascular ICUs at University of Michigan,

(00:21):
after doing an anesthesia critical care fellowship a
couple years ago.
And what we're gonna talk about today is
kind of the management of refractory dispute distributive
shock
and focus on one of the novel therapeutics
for that, which is angiotensin 2. It's called
GIAPREZA. You may have heard it referred to
that as the trade name. But,

(00:41):
keeping it best practices, I'll refer to it
as ANG2 or angiotensin 2 throughout the rest
of my talk. So just a quick word
about epidemiology and history. There's about 1,700,000
hospitalizations,
of patients with distributive shock. Overwhelmingly,
that's driven by sepsis or septic shock, which
is probably about 75 or 80%.

(01:02):
But a lot of other patients exhibit distributed
physiology. So if you work in cardiac surgery
ICUs, your patients coming out of the OR
will often have some distributed physiology from their
postponed vasoplegia.
If you work in trauma ICUs, you'll see
patients with neurogenic shock, the occasional anaphylactic shock.
They all,
exhibit this distributed physiology.

(01:23):
The mortality of distributive shock and particularly septic
shock hangs out at about 25%.
Now this increase is quite significantly when you
start getting additional organ dysfunction. So, for example,
if you go into dialysis,
it bumps up to 50%,
or if you're requiring
more than one vasopressor
after 24 hours, the mortality jumps up quite

(01:45):
significantly as well.
And,
one of the things to remember about this
is that it's not so much
getting a drug or the bacteria disinfecting you
or even the organ system
that is driving the mortality.
What drives the mortality is the number of
organ systems that are impaired
and how severe their impairment is. And you

(02:06):
can think about that with, for example, the
SOPA score. You get points for SOPA score,
for having neurologic dysfunction, cardiovascular dysfunction, renal dysfunction,
liver,
platelets.
But then as you move down the spectrum
of illness and require
more organ support, such as higher doses of
vasopressors,
more oxygen therapy,

(02:26):
the score goes up. And that's important to
remember when we're thinking about prognosis
and, you know, how good our therapies are
gonna be. It's one thing to change a
number,
such as a blood pressure. It's a total
different thing to change,
someone's mortality
or their organ dysfunction or their need for
organ replacement therapies.

(02:48):
So let's talk about a brief history of
the management because I I finished medical school
in the early 2000,
residency in 2006.
And back then,
early goal directed therapy was a thing, and
that was driving
a large amount of crystalloid resuscitation,
usually at least 20 cc's per kilo.

(03:08):
More like 30 or more, I would say,
in the emergency department in the first 24
hours.
And
we are starting to realize
over the past, I would say, 5 to
10 years that that one size fits all
strategy of fluid resuscitation
is not beneficial to every patient. And we
probably need a little more tailored approach that
some patients do better with more fluids, others

(03:30):
do better with less, and that probably a
one size fits all strategy,
is at least not all that effective and
maybe causing death by saltwater drowning.
Up until the early 2000, though, interestingly, norepinephrine
was really not considered to be a first
line vasopressor.
In the late nineties, when I was in
medical school in early 2000, dopamine was favored

(03:51):
at most institutions.
And we even had a saying when I
was a med student, leave them dead with
Levothy, Levothy being the trade name for norepinephrine.
And some of this was a fact that
norepinephrine
wasn't started until patients run very high doses
of dopamine.
And what we're gonna see with the study
that we conducted at Michigan,
when you start adding vasopressors

(04:12):
on the back end of patients,
1 or 2 days into it with a
lot of organ dysfunction, there's not a a
drug that's gonna fix that, and no vasopressor
is gonna do very well.
But over the, let's say, about the past
20 years, there's been a a move in
the literature starting around 2,006 to 2,010, I
would say.
Looking at norepinephrine, establishing it as the first

(04:33):
line vasopressor over dopamine.
And at first, the outcomes that were changing
were less arrhythmias with dopamine.
But then,
about 10 years ago, we started to get
enough signal of mortality that all the guidelines
from surviving sepsis and the pros were using
norepinephrine.
And dopamine had been kind of relegated

(04:53):
to use as an inotropic agent or a
chronotropic agent and largely out of medical ICUs
and more cardiac
or cardiovascular
ICUs.
Interestingly,
norapy is gonna be compared against epinephrine for,
septic shock against phenylephrine,
vasopressin,
and even terlipressin, which is available in Europe
and may be coming to the United States

(05:14):
soon. But there's not been any real big
study to show a meaningful
improvement of 1 vasopressor
over the other for distributive shock. But by
and large, what I would tell you is
that it's norepinephrine
is generally a a b plus,
vasopressor
for most conditions that need a vasopressor. It's
a great place to start, and rarely is
it ever the wrong vasopressor.

(05:36):
So I think, you know, for distributive shock,
particularly sepsis, norepinephrine is it, and that's what
we're using at every place I've worked at
for the past 15 years.
But what do you do when norepinephrine
doesn't seem to work? And,
you know, this talk, we're gonna talk a
lot about esoteric drugs that you probably don't
have access to. But this slide, if you're

(05:57):
a fellow or you're a resident,
I really want you to take home. Because
it's not so much the drug, it's your
mindset of approaching what to do with your
norepinephrine. Is it working? And this is really
important. Your first step is first ask yourself,
are you really treating distributive shock?
Because when you've done this enough, you will
get patients that were billed as, oh, they

(06:18):
just came out of the OR with a
stent in their ureter. They're gonna be a
little surzy from that, and no one told
you that they also have a massive retroperitoneal
bleed.
Or they have tamponade, or they have RV
failure, which if you've managed enough RV failure,
you'll see plenty of it masquerading as septic
shock.
And these patients,
don't do well with purely vasoconstrictive

(06:40):
therapies. They need blood volume. They need their
tamponade drain. They need their right heart support.
They need may need mechanical circulatory support.
So if your first line vasopressor
is reaching rocket fuel levels going up very,
very, very, very fast, One of the best
things you can do to save your patient's
life and prevent the downstream complications of multisystem

(07:02):
organ failure is to ask early,
am I on the right trail here, or
am I missing hypovolemia,
cardiogenic,
or obstructive shock,
and they need something other than just a
vasopressor.
The second step that I do is I
ask myself before I start reaching for a
second vasopressors, do they have some metabolic process

(07:24):
that needs hormone or vitamin replacement? So think
hydrocortisone
or thiamine
or thyroid or something like that.
Is this a patient that was on large
doses of steroids before they got sick, and
they now have relative adrenal insufficiency,
and we need to replace that? Or is
this an alcoholic, for example,

(07:44):
who, has depleted the thiamine stores, came in
with alcohol ketosis, and you tied those thiamine
because they're lactates 10?
Step 3 is then, yeah, if you've answered
those first two steps and you've convinced yourself
that it is distributive shock and you need
a second line vase vasopressor,
let's go ahead and get it done.
Give me an idea of what you're hoping

(08:05):
to do,
when you add a second line vasopressor.
Alright. So we're getting some votes here.
Improve organ dysfunction.
We had someone,
I think, vote for
reduce the duration of shock,
prevent ischemic events.
So it looks like improved organ dysfunction

(08:26):
is by and far away
winning this,
pulling out
of the gate.
We'll give it another 20 seconds or so.
Alright.
Let's let's talk about it, about what the
evidence is about,
adding additional vasopressor. I'm gonna go ahead and

(08:46):
move on from this.
So
why are they needed? So, interestingly, catecholamine vasopressors,
they're associated with a lot of arrhythmias.
If you spend enough time on a catecholamine,
such as norepinephrine and epinephrine,
you're gonna, about 50% of the time, get,
some arrhythmias,
overwhelmingly

(09:07):
atrial in nature, AFib.
Myocardial ischemia,
and we define that as a small bump
in your troponin,
all the way ranging up to
frank SD segment changes in SD segment elevation
is a much smaller percentage at about 15%.
Postoperative
cardiac surgery patients are particularly prone to these

(09:27):
arrhythmias, and many cardiac surgeons,
if you ever work in a CVICU
and cardiologist and and cardiac ICUs,
will limit the amount of catecholamines
due to worse outcomes associated with these atrialythmia.
So,
in our CVICU at University of Michigan, you
could not go up more than 0.2 mics
per kilo per minute of norepinephrine

(09:47):
without getting the surgeon on the phone and
talking with them about it.
And, again, most of these arrhythmias are atrial.
And when you look at atrial arrhythmias in
the ICU,
they are associated with an increase in mortality.
But it's unclear to us if it's actually
the atrial arrhythmia that's driving that mortality, the
AFib is, or are the things that kill

(10:08):
people also cause AFib before they kill people.
And,
when we look at vasopressin, it's a non
catecholamine,
and the only real benefit,
to it
is in preventing actually arrhythmias,
and other adverse events related to the arrhythmias
that are related to catecholamine exposure. It hasn't

(10:30):
really been shown to change mortality
or SOPA scores, which looks at organ dysfunction,
but it's really reducing your arrhythmia risk,
is what the best evidence
for adding a noncatecholamine
vasopressor is.
It doesn't also improve ischemic events. So,
vasopressin,
when used at higher doses, is a very

(10:51):
potent potent splanchnic and limb bay vasoconstrictor
and can easily
precipitate,
mesenteric ischemia and limb ischemia
at its higher doses.
Let's do another poll real quick. I'm a
little bit curious, and I'm not sure if
at University of Maryland, if you use weight

(11:11):
based dosing,
or just microgram dosing. So I put this
in micrograms
per kilogram per minute or a rough,
microgram per minute.
I wanna get an idea of where people
feel like when they get to a dose
of norepinephrine,
they need feel the need to start adding
in an additional vasopressor.

(11:33):
So I'm seeing a lot of votes between
that 0.2 mic per kilogram per minute to
0.3 right now.
That's gonna put you at about 15 mics
per minute and 20 mics per minute. And
let me just say, I I've trained under
the weight based dosing,
and I practice now under
just the, you know, numerical dosing, not weight

(11:53):
based. And I vastly prefer weight based dosing
when I'm dosing vasopressals because I I find
that particularly when I'm dealing with trainees,
they have a number in their head, and
that number
in a 50 kilo,
elderly person is vastly different than that number
in a
300, you know, pound,
Baltimore Ravens, you know, lineman.

(12:16):
But it looks like most of you are
hanging out in that point 2 to point
3 range. And that's that's where
I am for the most part is in
that range.
So
most of the guidelines, for example, surviving sepsis
will say
neuroepidosis
of 0.15 to 0.3. Most settle around 0.2
to 0.3 mics per kilo per minute. There

(12:37):
are some things that will cause me to
go lower or even maybe even lead
with, vasopressin as a first line agent very
rarely. So if I'm dealing, for example, with
a bad RV,
vasopressin
does not cause the pulmonary vasoconstriction, at least
we don't think it does.
That catecholamines and particularly,

(12:57):
certain non catecholamines such as phenylephrine
may cause. Or if I'm dealing with a
lot of atrial arrhythmias and I wanna spend
some money on vasopressin and phenylephrine,
I might go with lead with vasopressin first.
Also, if I'm dealing with someone who's brain
dead or approaching brain death and they had
bad central DI,

(13:18):
I may lead with vasopressin because you do
get,
some water sparing effect that with that is
about, I guess, a third or so of
what you're gonna get from d d a
v p. So those are all reasons why
I might start
vasopressin at a lower dose.
One thing to remember about vasopressin is that
at the dose that we frequently see it
started at, which is around 0.03 units per

(13:40):
minute,
you're essentially replacing
a relative deficiency of vasopressin,
particularly in septic shock states that your body
is is not producing in that stress state.
Now you can increase that dose, and I
in my current shop, I find a lot
of people reluctant to go above 0.04.
But, in cardiac ICUs where we limit catecholamines,

(14:03):
we would go up to
0.083,
which is like, if you're dosing it in
units per hours, 4 to 6 units per
hour.
And at those doses, you're getting a very
potent vasopressor
effect
by binding to these v one receptors,
and you get all the concomitant then risk
of mesenteric and limb ischemia and even arrhythmias

(14:25):
interestingly at these very high doses due to
potentiation of sympathetic activity.
But, generally, at these lower doses,
it reduces the AFib risk, the relative risk
by about 0.77,
95% confidence, and on the meta analysis was
less than 1. But no really other improvement
in other outcomes.

(14:45):
How people wean the vasopressin is very variable.
I personally
peel it off first as the norepinephrine doses
are coming down. So let's say we went
up to point 4, point 5 mics per
kilo per minute of norepinephrine.
I then once I get down to point
2 or so, start asking the, nurses to
come down on the vasopressin,

(15:06):
particularly if they were higher than point 3.
I usually ask them just to go down,
you know, by point o one every 30
minutes until it's off,
when I'm doing that as opposed to going
down to point o three and just turning
it off. But some people will go down
to point o three and just have them
turn it off at that point. I think
that's fine. And the reason why I wean
the vasopressin first is it's typically more expensive,

(15:31):
than norepinephrine.
And in our shop a year ago, it
was 500 to $700
for a bag of vasopressin.
That price may be coming down, I think,
either now or in the near future, has
come down as it goes off patent,
but it's still more expensive than norepinephrine
and less titratables.
It takes longer to come off.

(15:52):
What about
angiotensin 2 as a second line presser? Well,
it activates your RAS system,
and it has a lot of effects. You
have multiple types of angiotensin
receptors,
that involved in the vasomotor response, such as
the vasopressin 1 receptor,
your sodium and water reabsorption,

(16:14):
inflammation, apoptosis,
thrombosis. So it has a lot of off
vascular effects.
But we've known since the sixties that it
was a potent
vasoconstrictor.
And there was a lot of animal animal
data using bovine,
for example, angiotensin
2,
as a vasopressor
in, basic science labs and animal models.

(16:37):
The ATHOS,
trial was a pilot trial that used recombinant
human angiotensin 2,
to show a reduction in norepinephrine
doses
in a small number of patients. Tim got
the drug, Tim got placebo
at a fixed dose, and it basically just
showed that we were able to reduce the
norepinephrine,

(16:58):
by giving people angiotensin 2 compared to receive
placebo. And this, again, was just, you know,
phase 1 pilot type data.
Not a big end, certainly not practice changing.
But I think if we're gonna talk about
the study we did, we need to have
a little bit of background
with the larger study that came on after
that, which is APHOS 3.

(17:19):
So this was an industry sponsored phase 3
trial to determine if the addition of angiotensin
2 to standard care, which would, at that
time, be largely norepinephrine and vasopressin,
would improve would improve blood pressure. It was
power to improve blood pressure, not to change
mortality
in patients with distributive shock.
It was blinded and randomized, and it took

(17:41):
place in about 75
ICUs across North America, Australia, Asia, and Europe,
from about 2015 until 2017.
Now,
there are a couple things that we need
to understand about,
this trial and its methods
and also about,
a term that you will hear that I'll

(18:02):
throw out in this pay in this talk
called norepinephrine equivalent doses.
So let's talk about norepinephrine equivalent doses first.
If you use norepinephrine
as a standard,
we'll say point 1 mics per kilo per
minute,
you can take other drugs such as epinephrine,
dopamine, phenylephrine, and vasopressin,

(18:22):
and convert them to a rough norepinephrine
equivalent dose.
And this is largely based on animal data,
and how true it is
is probably anyone's guess.
And this figure
appears in the supplemental
material,
in APHOS 3, and we used it in

(18:42):
our trial as well.
And it has been battered around,
the Internet and the literature for quite some
time.
One of the things I wanna draw your
attention to on this chart to the right
is phenylephrine.
So phenylephrine
is a fairly weak
vasopressor
compared to norepinephrine.

(19:02):
And in my current shot,
when we get to higher doses of norepinephrine,
I see a lot of people adding phenylephrine
onto that. I'm not quite sure why that
happens because when you're on rocket fuel doses
of norepinephrine, it's hard for me to imagine
that at its potency, phenylephrine is adding much
to the alpha effects that norepinephrine
isn't already hitting.

(19:24):
But I just wanted to point that out.
And in APHOS 3, what they did is
they took patients with distributive shock,
and they define that
as patients with shock that had received 25
cc's per kilo.
And they had to have certain physiology
that was pretty rigorous to prove it was
distributive and not hypovolemic,

(19:44):
or cardiogenic. They had to have
either an SCVO 2 greater than 70%
and a CVP greater than 8 or a
measured cardiac index,
more than 2.3 liters per minute body surface
area.
Then they excluded,
a lot of patients that, you know, I
would expect them to exclude. It would be

(20:05):
burns, trauma,
patients on ECMO,
cancer patients with neutropenia.
And what they came up with with this
study was about 400 eligible patients.
344 were randomized after consent,
to a 172 placebo, a 172 got angiotensin
2. And then they had a few number
of withdrawals that, in the study

(20:28):
that they continued and used a modified attention
to treat analysis.
So in the protocol, which is kinda tricky,
and I think you gotta have a good
understanding of this protocol in order to evaluate
this study in your scope of practice,
is that the treatment arm, the patients who
are randomized to receive angiotensin
2

(20:49):
got the drug started at 20 nanograms per
kilogram per minute. And then this dose was
adjusted over the first three hours
to increase the map to at least 75
while holding constant
the background dose of the other vasopressors that
have been started.
And I believe to get into this trial,
you already had to have been on 0.2

(21:09):
mics per kilo per minute of norepinephrine
equivalents.
So that's either just norepinephrine or largely norepinephrine
plus the vasopressin.
And then after this 3 hour period, the
dose of angiotensin
2 could be adjusted
between 1.2540
nanograms per kilogram per minute.
Again, all while keeping

(21:31):
the background dose of the first line or
second line vasopressors
constant or lower.
And the control arm was receiving a placebo.
So as you can imagine, the placebo was
doing nothing.
And for safety,
the patients would have to have if the
placebo is doing nothing, the background dose of
their norepinephrine and vasopressin would largely have to

(21:54):
be adjusted or go up if their shock
or hypotension is worsening.
And non responders
were those who got an escalation
of the background of vasopressors.
So, again, they weren't looking so much at
mortality as they were a blood pressure response
to receiving angiotensin
2 and what they consider to be a
blood pressure response was

(22:15):
maintaining a map
and being able to prevent the escalation and
preferably the reduction in doses of background vasopressors.
So let's take a look at their patients.
They were older. The mean age is about
65.
Most of them were septic, about 75%,
and they were pretty sick. So the SOPA
score on average at enrollment was around 12,

(22:38):
which would correspond to more than 95%
predicted mortality.
ARDS was in about a third of their
patients,
and 70% of them had received vasopressin. So
they were on norepinephrine
plus vasopressin, 70% of them,
at enrollment.
And
their
average norepinephrine

(22:58):
equivalents was about 0.34,
which is pretty high.
15% of their patients
had received an angiotensin
receptor blocker or an ACE inhibitor at the
time of enrollment, and that is a particular
interest patient population because this drug
should be
acting on those receptors.

(23:20):
So their primary endpoint
was the MAP response at 3 hours.
And they showed a fairly convincing elevation in
blood pressure
at the 3 hour mark in the patients
receiving angiotensin 2 and placebo.
The secondary outcomes is where it gets a
little tricky
and I would say a little bit problematic

(23:42):
for me at least.
They noted that the cardiovascular
SOPA score also improved.
The reason why this is tricky, if you
remember from the SOPA score, is that you
get points
for having hypotension
and the amount of catecholamine
vasopressors
that you're on.
The higher the amount of,

(24:02):
norepinephrine
or dopamine that you're on, the more points
you get.
But there is no,
allotment in SOWFA for angiotensin
2.
So by going up on or adding angiotensin
2, yes. You you raise the blood pressure
by adding angiotensin 2, and you lower the
amount of norepinephrine you went on. But in

(24:24):
reality,
angiotensin 2 is just a different vasopressor.
And
the is it really a a should it
be a
reduction in your cardiovascular SOPA score when you
substituted out one vasopressor for another that doesn't
happen to be in the scoring system?

(24:44):
The overall change in SOPA scores
looking at middle status, liver dysfunction, kidney dysfunction,
lung
dysfunction
didn't change.
So that would support
my theory that all they did was switch
vasopressors around. They switched from a catecholamine,
norepinephrine,
and a little bit of noncatechol

(25:04):
vasopressin
to,
a mixture of
catecholamine and norepinephrine, vasopressin, and now angiotensin 2.
Other
endpoints that were more exploratory and secondary were,
mortality at 7 28 days.
There was a trend towards improvement in the
patients that got angiotensin 2, but it did

(25:26):
not meet statistical significance.
So what we came away from this paper
saying was is that,
yes. Angiotensin 2, there's good evidence that it
changes blood pressure. There's not convincing evidence that
it meaningfully changes your organ dysfunction,
and there's not convincing evidence that it's gonna
change your mortality.

(25:49):
These are two figures
that came from the
APHOS 3 trial, and you can see
the mean arterial pressure over time,
in hours since the start of infusion. In
that first 3 hour period where they added
the angiotensin
2, you see this quick spike up in
the blood pressure in the patients who received
angiotensin
2. And then

(26:10):
in the titration phase of this trial,
that blood pressure comes down and largely stays
above,
the standard treatment arm all the way out
to the end of the study.
And then we see the baseline doses of
vasopressors
did improve for both,
sets of patients, both placebo
and angiotensin

(26:30):
2,
but that change occurred a little bit better
or faster, I I guess you could say,
in the patients receiving angiotensin 2 compared to
those who received
placebo. What about adverse events? So there's overall
similar rates of adverse events, and that was
quite high. Keep in mind that when you

(26:52):
do these trials,
a vigorous sneeze is an adverse event.
So,
you know, you look at this, you say,
oh my gosh. 90% adverse events.
That that's a protection,
done to guard patients in these trials.
Serious things like Vtach and and afib
and limb ischemia, intestinal ischemia, stroke, and all

(27:14):
that were actually quite low and similar across
the board between the placebo and the patients
who received angiotensin 2.
In the trial, they reported 3 DVTs in
the a t two group and 0 in
the placebo.
There was also a similar pattern of increased
antithrombotic
medication used in the angiotensin

(27:35):
2 group, and this led
after this trial, the FDA to require that
the manufacturer
mention
an increased risk of thrombotic events, because there
was also
a signal for some arterial thrombotic events was
my understanding.
And this, again, makes sense if you remember
from my earlier slide.
8 angiotensin 2 has a lot of off

(27:56):
a t one receptor effects
that are involved in thrombin formation, and it
actually increases thrombin formation. So,
uptick in thrombotic events was not too surprising.
There was no difference in the proportion of
patients developing
infections
in this study,
but
more of the a t two patients developed

(28:17):
multiple infections
or fungal infections, which are a lot should
be a lot more uncommon, of course, than
bacterial infections.
And there was a question of maybe this
might be compounded by ACE inhibitor use. So
when you're given this drug, you may not
be getting a free ride when comes to
your clots or your infections,
and that has been an area of interest

(28:38):
for further study
after this trial is published.
So what are some of the problems with
this trial? Well, 1, this trial
looked at a what I call a monitor
associated outcome, which is blood pressure.
So, you you know, you have your moves,
which are your monitor oriented outcomes,
your poos, your patient oriented outcomes,

(29:00):
your dues, your disease oriented outcomes.
And, you know, I mean, changing someone's blood
pressure is nice, but
changing the duration of their shock, their mortality,
their organ dysfunction, need for organ replacement therapies
is is far more meaningful to me.
Again, the method that a t two is
applied
where they added the a t two and

(29:22):
then weaned the first line vasopressors
is not realistic when the drug costs more
than $1,000
per vial. I don't think any institution is
gonna allow you to do that,
to weed off the norepinephrine
while leaving the very expensive drug running.
Another thing is that to get into this
trial, you had to be on 0.2 mics

(29:42):
per kilogram per minute, and they call that
high dose vasopressors.
Maybe point 5 mics per kilo per minute
might be a better definition
of high dose vasopressors.
That's again a an area of debate for
people. And I already mentioned the change in
SOPA score in terms of the cardiovascular component.

(30:03):
That's a bit misleading because
you just don't get points for bay for
angiotensin too because it wasn't around when the
SOPA score was created.
And by artificially lowering the amount of norepinephrine,
does it really improve someone's cardiovascular
performance just because you change it to a
different vasoconstrictive
agent?
We don't really know how the use of

(30:24):
steroids,
impacted APHOS 3 since adrenal and approach
had not yet been published.
But what we were left with this trial
was that we had pretty high and, I
mean, me, I think the the critical care
community,
pretty high confidence
that a t two improved blood pressure,
but we didn't really see

(30:45):
a a meaningful improvement in a outcome like
mortality or ICU length of stay,
duration or shock, or need for organ replacement
therapies. So there wasn't a big push for
us to use it unless
we were in a salvage situation. We were
throwing the kitchen sink at patients to try
to save their life.
So this period from the time the study

(31:07):
was published from
2017 to 2282, as you can imagine, not
much was happening in the world of critical
care.
I say that kind of facetiously
with COVID happening. You know? I mean, we
had a lot of sick patients, a lot
of opportunities to use this drug.
But the use of a t two had
become highly variable across institutions

(31:27):
with a lot of institutions kind of looking
at it like drugs like indexinet
and not even stocking it because it was
expensive, and they just didn't want people using
a drug that hadn't been shown to improve
mortality.
There were also a lot of poke poke
stock analysis of APHOS 3 from,
Ashish and his group,
suggesting when they looked at their data again

(31:49):
that, outcomes might improve in patients with high
renin
or those that have received kidney replacement therapies.
Analyzing renin back then was kind of like
doing,
cosyntropin
tests. They're cumbersome.
They're difficult. They're time consuming. They give you
a headache just thinking about it generally.

(32:10):
So not a lot of places were guiding
their use of angiotensin 2,
with renin.
There was also data looking at angiotensin
2 in certain patient populations, such as the
post cardiac surgery patient population
showing improved outcomes.
And a lot there are several studies looking
at small numbers of patients when it was
used as a salvage or 3rd or 4th

(32:31):
line vasopressor,
showing that maybe it improved blood pressure, but
none of those were ever powered to to
show a change in mortality.
So what were we doing at University of
Michigan between
2,008
or 2018 and 2021,
we did our study?
Well, norepinephrine was overwhelming the first line vasopressor

(32:53):
that we used across our ICUs for distributive
shock. Like I said, we did limit the
dose to point 2 mics per kilogram per
minute in our cardiac and CV ICUs,
but the sky was the limit in the
medical ICU, the neuro ICU, the COVID ICU.
Vasopressin
was,
by far and away, our second line agent.

(33:15):
There was a best practices advisory. We used
EPIC that would fire EPIC.
If you tried to start vasopressin and your
neuropathy dose was less than 0.15,
it would advise you that what you were
doing was not very evidence based, unless
there was a competing condition such as right
heart failure
or or arrhythmias.
It didn't restrict your use. You just had

(33:36):
to acknowledge
the BPA and why you were why you
were doing it if you were starting in
a lower dose.
We used a fair amount of epinephrine, but
that was usually used in inotropic doses of
less than 0.05 mcgs per kilogram per minute.
We rarely used it as a vasopressor.
Even in post cardiac arrest situations,
patients would come to us from the ED

(33:57):
on epi at vasopressor doses, and we quickly
try to get that down to inotropic doses
and switch over to norepinephrine.
Phenylodopa
were more uncommonly used. Phenyl, we typically use,
post anesthesia, just counteracting some anesthetic,
that we thought would come off once the
patient's extubated
or if there was an arrhythmia risk. And

(34:18):
dopamine, we had some cardiac surgeons who were
just old school, and they use dopamine as
an hydrotrope largely.
At 2 was introduced to our formulary back
in 2018. There was no requirement that any
clinician
document distributed physiology
using a CVP
or a central venous oxygen saturation or cardiac
index. You just you could order it.

(34:40):
And the worst that would happen is the
pharmacist
would call you and ask you, do you
really wanna do this? But that was not
a a consistent thing.
We used it disproportionately
in our CVICU
likely because of that need to limit the
amount of catecholamines.
And then in that era, what we saw
was an increasing use of steroid or sepsis

(35:02):
dose steroids
at 200 milligrams
of,
hydrocortisone
a day, either as an infusion or divided
q 6 to q 12.
Based on the adrenal and the approach trial,
then based upon the emergence of COVID,
our group embraced steroids fairly early for COVID.

(35:25):
We then decided since we had all this
data, it was time to do a study.
We had reached about 271
patients, which is a pretty robust spread across
all of our adult ICUs, which is at
least 8 ICUs,
that treat adult patients at the University of
Michigan.
We initially wanted to do this as a
propensity
match trial where,

(35:46):
we match patients based on dose of medication
plus other factors such as,
you know, if they had cancer or if
they had cardiac surgery.
And that's how propensity score matching
is done in order to
eliminate
baseline variability,
and equalize the patient populations, but the problem
to do that is we had just so

(36:06):
much data.
We had thousands of data points on individual
doses of drugs,
and to try to,
to try to
to do propensity matching on other factors
would have required so much computing power. I
probably would have risked SkyNet coming alive and
becoming self aware
and bringing about judgment day or something. And,

(36:27):
like, there was not a computer I had
access to that could do this. So we
had to settle on, and that was a
little bit of an eighties reference for you,
Terminator buffs.
But,
to settle on just matching on the norepinephrine
doses, and then on the back end,
adjusting using,
regression analysis and multivariate analysis
in order to equal equalize the patient populations.

(36:50):
But what we did was we matched patients
on the norepinephrine
equivalents that they received,
when the a t two was started.
And we decided to,
match them on a concurrent control patients
from 2018
forward when,
angiotensin 2 is available,
but who did not get angiotensin 2, but,

(37:11):
also, on a separate
historical control,
to account for changes in behavior that might
have occurred
with angiotensin 2 arriving.
So we had both the concurrent and the
historical control.
And what we did was that we monitored
patients for 5 days after the angiotensin 2
was started,
looking at

(37:32):
organ function, SOPA score,
largely, blood pressure response. And then, we looked
at organ failure and mortality
out to 90 days. And by organ failure,
I mean, need for kidney replacement therapies,
need for mechanical ventilation
out to 30 90 days for mortality.
We use APHOS 3 to do a power

(37:54):
calculation.
There, they found an absolute reduction in mortality
of 8% that not
did not meet statistical significance.
We estimated that we would need about
a 198
angiotensin
2 patients,
to provide us with 90%
power,
using our 2 control method,
and we easily surpass that,

(38:17):
power. And ours would be the only study
that would be calculated,
power to detect the mortality difference based on
what EPOS 3 had seen.
So what do we see with our results?
So much like ATLAS 3, the overwhelming,
patient that got this drug were patients,
who are deemed to be septic,

(38:37):
and we used the traditional definition of sepsis,
I e,
they had organ dysfunction plus an infectious organism
and received antibiotics.
That was about 75 to 80% of our
patients.
We were a little bit younger,
than APHOS 3 with 57%
of our patients. Excuse me. 8 minute

(38:58):
mean mean age was 57.
We also had enrollment when the a t
two was started was at a little bit
higher dose. You'll recall
in APOS
3, the norepinephrine equivalence is closer to 0.3.
We were at 0.43
mics per kilo per minute,
of norepinephrine.
And most of our patients were on more

(39:18):
than 2 vasopressors when it was started. So
we were really using it by and large
as a salvage vasopressor
when at least 2, if not 3 agents,
was not working.
Our mean SOPA score was 12,
versus 10, so there was a little baseline
imbalance in the angiotensin
2 group that we had to account for
on the back end. That means that patients

(39:39):
were using a t 2 in a sicker
patient population.
But there was no imbalances when it came
to other things that are used for shocks
such as steroids, methylene blue, and vitamin b
12.
This slide
shows you
the distribution
of norepinephrine
equivalent doses by number of patients in the

(40:00):
y axis. And when you look at it,
you're gonna say, wow. Those graphs look identical
in a, b, and c, a being,
the angiotensin 2,
b being the concurrent controls, and c being
the historical controls. They should look identical because
the patients are matched. So for every patient
in an a t in an angiotensin 2

(40:22):
group, a patient getting an identical amount of
norepinephrine,
got it in both the concurrent and in
the historical control groups.
The purpose of these figures is so you
can see where most patients fell
in terms of getting norepinephrine
equivalents. It was in this range of
point, you know, 1 to point
5 to 6 mics per kilo per minute

(40:43):
of norepinephrine. But, of course, we had some
patients
getting more than 1.5
mics per kilo per minute worth of vasopressors
and norepinephrine
equivalents.
And then down in figure d, you see
the time course, the duration. Most of our
angiotensin 2 was given for about half a
day,
and then it petered off out to 5

(41:05):
days. There were some patients who were just
hangers on, the ones with just refractory shock
that the angiotensin 2 couldn't be pulled off
of.
So what were our outcomes? Our primary outcome
was mortality at 30 days. We saw no,
difference between patients
in the
concurrent control, historical control, or we aggregated it

(41:26):
together,
the controls into 1 large control patient population,
no difference in mortality
whether or not they had got angiotensin
2. There was also no difference in mortality
at 90 days, which is a sec secondary
outcome.
Figures a and b over here look at
change in SOPA score in the first five

(41:47):
days
after the angiotensin 2 was given,
and we see that,
no real difference between those receiving 82 and
control.
Their organ their SOPA scores went down and
proved
over the 5 days, regardless of whether or
not they got the drug. And much like
they did in the APHOS 3 trial, we

(42:07):
reproduced a figure
of blood pressure,
and we really didn't see
a big difference change in blood pressure in
these patients either, whether or not they got
angiotensin
2 versus control. And, again, that's probably because,
we were weaning the angiotensin off pretty quickly
if their blood pressure was responding as opposed
to leaving it on and weaning off

(42:28):
the first and second line, vasopressor agents.
So what did we find were the drivers
of mortality? Well, kind of the same thing
I that we found in on just about
every study looking at sepsis,
over the past 30 years, and that is
that most things such as organ dysfunction,
higher lactate, higher doses of equivalent,

(42:50):
more need for renal replacement therapy, those are
the things that that determine whether or not
you were gonna survive.
Angiotensin 2 did not impact,
the need for kidney replacement therapy,
which is what KRT stands for. That apparently
is the new hotness in terms of,
we don't call it RRT anymore in the
research world. We call it kidney replacement therapy,

(43:12):
not renal.
And no change in mechanical ventilation as well
or thrombotic events. Now in our trial, thrombotic
events were high,
probably because we were in a sicker patient
population than the at those three trial, but
it was uniform across all three patient populations.
So what were the strengths and the weaknesses

(43:32):
of our trial? So strengths were powered in
insect mortality. We had a diverse patient population
across multiple ICUs.
This wasn't just done in medical ICUs or
neuro ICUs. It was medical neuro, cardiovascular,
and cardiac in adults.
We did not enroll patients out of the
operating rooms.
We also found that,

(43:55):
we well, the reviewers required that we do
a very detailed sub patient population analysis
looking only at patients
with clear cut sepsis.
And, again, this trend of no change in
outcomes
held up in that patient population as well,
and that was available,
in our supplemental data.

(44:16):
The other thing is that this is a
realistic look. You know? You hear about
efficacy versus effectiveness,
and this looks at effectiveness
outside of the confines of a rigorous,
clinical trial where you have an IRB
evaluating everything you do. This is what happens
when you unleash this drug on the world.

(44:36):
And then we also looked at a large
number of patient oriented outcomes, mortality, organ dysfunction,
not just at blood pressure.
The downside is this is retrospective data. It's
a single center, so Michigan may not be
the same as Maryland, may not be the
same as Atrium.
We clearly had an imbalance,
and then there's not a way to account

(44:57):
for every
confounder,
although
our multivariate
analysis on the back end, I thought, was
quite thorough.
But there might have been other confounders that
we couldn't account for, and, actually, trying to
propensity
match that on the front end would not
have helped that. There's really not a big,
improvement in finding those confounders when you go

(45:19):
try to do it on the front end
with propensity score matching as opposed to on
the back end with your multivariate analysis.
We occasionally found conflicts in the chart on
sharp shock type and discordant data.
For example,
someone might shoot a cardiac index, a thermodilution
cardiac index at the same time as they

(45:39):
obtain,
central venous oxygen saturation or mixed venous oxygen
saturation.
And one is pointing one way and other,
the the index is pointing the other. We
largely went with the cardiac index as a
numerical as opposed to a fit calculated in
those circumstances,
or,
they had both distributed
and

(45:59):
hypovolemic
physiology. And that just happens sometimes. You get
these patients
with terrible hemorrhagic shock or terrible cardiogenic shock
that then, after
hours, develops a vasodilatory
component due to the inflammatory
cascade, and they're gonna need some vasopressor. You
can't just support them with blood transfusions.
So,

(46:21):
it all depends upon, you know, how far
you go down this shock paradigm, this rabbit
hole. It often doesn't fit one one paradigm.
Distributive,
becomes hypovolemic, and hypovolemic becomes some distributive
once someone has had been on high dose
vaso compressors for more than a day.
And then we were not powered to find

(46:41):
subpopulations
that might benefit from a t 2. And
the reviewers asked us, you know, do we
check renin? You know, we we didn't check
renin. We checked a lot of lactates, but
no renin.
And we don't know what happened in early
shock because, again,
most of our patients were on 0.4 mics
per kilo per minute before anyone gave them
angiotensin
2.

(47:02):
So what are the implications?
What I take from this is that adding
angiotensin 2 or, I would say, any other
vasopressor right now
as a 3rd or 4th blood agent
is not been shown to improve outcomes yet.
So my practice currently is to just increase
the catecholamines

(47:23):
and the, vasopressin as local
customs allow. And I personally
no longer use,
angiotensin
2 for refractory distributive shock when I'm allowed
to increase the dose of my catechols.
Now if I'm in a heart transplant or
post cardiac surgery,
patient population where they're artificially limiting the amounts

(47:44):
of catechols or have a patient with bad
RV failure,
I personally believe angiotensin 2 probably doesn't basically
constrict your pulmonary vascular quite as much. I
will use it, and I'll use it earlier.
I still think that there is an echo
points regarding angiotensin
2's role as a first or second line
vasopressor.
And there actually is a trial looking at

(48:05):
postpartic surgery patients at a t 2 versus
nor epi
as a first line vasopressor
or in patients receiving
angiotensin
blockers and ACE inhibitors.
And there's also needs to be some further
studies done
looking at a molecular phenotype. And,
Ashish Khan and I put in for an
institutional grant,

(48:27):
measuring renin and prorinin
is no longer as cumbersome as it used
to be. In fact, there are now point
of care testing that you can do. So
we're trying to develop
some, renin angiotensin,
molecular phenotypes for patients
at various stages of shock to see if
we can target maybe a t 2 to
certain patient populations.

(48:49):
But that's all
exploratory at this point.
Briefly, let's talk about other salvage vasopressors that
I'm seeing used a lot. So methylene blue
and b 12 are being used,
now.
They first got popular, I would say, about
10 or 15 years ago,
in the post pump vasoplegia patient's cardiac surgery.

(49:12):
We're now seeing them used in septic shock
as well. Both work through
inhibiting nitric oxide, which is thought to be
a major driver of vasodilatation and sepsis.
My, methylene blue also inhibits cyclic GMP formation,
which is a more downstream mechanism.
There've been some small studies looking at this,

(49:34):
and I would say the preponderance of the
evidence is that much like angiotensin
2, they do increase your blood pressure. They
are also on the expensive side,
but larger studies are needed to know if
they really change outcomes,
when added as salvage to patients on more
than 0.4 to 0.5 mics per kilo per

(49:55):
minute.
The other thing I think you need to
know about these 2 drugs if you're gonna
use them is that methylene blue does not
play well in the sandbox,
with people who are g 6PD deficient. So
don't give it to them, and we'll make
them worse probably.
And vitamin b 12 is, I believe, is
a chromophore, and every time I've given it
to patients on ECMO,

(50:16):
your,
return line
spectrophotometric
measurements of saturation start going haywire,
because it will it will mock with that
and mess with it.
So just be aware of that if you're
gonna give those drugs,
in those,
circumstances.
So that's all I got.
We've got a couple minutes, 5 minutes for

(50:36):
questions.
I wanna thank you for inviting me.
I really enjoyed,
talking to you and talking about this topic.
I I like the city of Baltimore a
lot. I wish good things for you and
your football team. We have a Harbaugh
roots.
So,
go Ravens. Go Michigan.

(50:56):
How can I answer your questions?