Resistant Hypertension – Trials and Tribulations

Resistant Hypertension – Trials and Tribulations


Thank you very much. Thank you very much
for inviting me today. This is going to be a slightly
different 45 minutes. We are going to discuss…. We are going to
discuss a case, but we also, I want to
encourage a discussion about how as clinicians we
interact with evidence in clinical trials, and a lot
of the discussion will be around trial data that came
out over the last five years or so, and our
interaction with that. So, this is where I work in London University College
in London Hospital, and here we run a tertiary
hypertension clinic for the surrounding area. Just to start,
let’s discuss a case. One of our patients that
was referred five years ago in 2012 by her
general practitioner. So, she had a background of
hypertension for many years, as well as many other issues. She had a meningioma in 2008. During the operation to
excise that meningioma, she had an episode of VT,
which means that she had an ICD inserted thereafter,
and luckily has had no further episodes of VT
since that was put in. There’s a questionable
diagnosis of epilepsy. The neurologists aren’t sure. She, again, had a single episode
but ended up on treatment. She has Type-2 diabetes,
but only on tablets, not on any insulin. So when she was
sent to our clinic, she was on five
different medications. I don’t know if anyone at
this very early stage wants to comment on the
combination of drugs that she was on. This was all managed by the GP. So she’s on both
an ACE and an ARB, as well as calcium channel
blocker and a diuretic, and a beta blocker. So she’s on five agents. She’s also on metformin
and sodium valproate. So in clinic, after
the history was taken, it was found that
actually she’s pretty well. She gets the
occasional headache, and she has no significant family history of
high blood pressure. She’s slim, got a BMI
in the normal range, and as per the GP
referral letter, in clinic, after
repeated measures, she really did have
very high blood pressure, so 195/110. But there was very little
else to find on cardiac examination. There was a
description of Grade 2 hypertensive retinopathy. Now, on these stages,
I’m sure many of you in this room do,
we undertook some baseline
investigations initially. Again, these did not
reveal anything untoward. She had a normal blood count. Her renal function was normal. Very mild proteinuria. Surprisingly, her
echocardiogram did not show any evidence of LVH, and that was also
true of her ECG. Now, what would you do
next working up patients with resistant hypertension? We routinely would perform
ABPM when we were doing this five years ago to have a
look whether patients have white coat hypertension, and
she most definitely did not. Her blood pressure was
sky-high at this stage. Now, that’s a mean blood
pressure, 190. She did have a low blood
pressure at one stage of 118/67, but you’ll note
that over 90% of her blood pressures were in the
poorly controlled range. So here we have
somebody with a diagnosis of resistant hypertension. She far exceeds the definition
in the European guidelines. She’s on five medications,
including the diuretic, and her blood
pressure is sustained. Professor [Staessen]
wants to advise you. Yes. I think this definition of
resistant hypertension is a very loose one, and we
should really leave it. I think without having checked
adherence of your patient, you cannot talk about
resistant hypertension. It’s as simple as that. You’re right. So, that’s exactly the
way we approach this. We approach this by saying
this person appears to have resistant hypertension, and
I’m sure it’s the experience of many people here this may be pseudo-resistant hypertension. So as I’ve put there,
she is failing to adhere to prescribed treatment. Sorry, could I just ask,
did you show us her nocturnal blood pressure? Did she have dipping? No. She didn’t dip to
a normal degree, so she still had a degree
of nocturnal hypertension as well as sustained
hypertension, both 24-hour and daytime. I am from Poland
[foreign language]. I think that’s the
echo in this lady, it doesn’t describe
any hypertrophy. If she really have
resistant hypertension, she should have
hypertensive left ventricle. So, for sure, she is a
failure adherence of her particular regimen. I think I normally
would agree with you. I think this lady is
quite unique in that, even to date, so I can give
you the up-to-date echo that was done a few months ago. She still doesn’t have
any evidence of LVH on echo, which is surprising when I
show you the results of her other tests. Similarly, the lack of
renal impairment’s very surprising as well. I don’t have a good
explanation for that, so there’s no big reveal
about why she has no real evidence of an organ
dysfunction at this stage. So we all tend to think of
at least two major causes of this phenomena. So first of all, is there an
underlying cause? We’ve just heard about FMD. Does she fail to
adhere to her regimen? If it’s neither of those,
then we can diagnose her with primary
resistant hypertension. So obviously, the first
thing that we do, well we do all
this in parallel, but of course we did
a secondary screen. Now a secondary screen in
my institution at the very minimum, after history
and examination, would include a look at
the amino arteries and the adrenal glands. Because of her ICD, this
was done with CT rather than MR. It did not reveal any
evidence of renal R2 stenosis, and I can tell you this
has also been and remains negative
on a second look. I would think that
if someone takes 40 mg Enalapril and
has severe hypertension, whether you find renal
artery stenosis there or not, will in no way be helpful as
a response to ACE inhibitors is perhaps the best
predictive measure of revascularization so
in my humble opinion, there’s no point in looking
into adherence makes much more sense than
looking into renin aldo. I’ll show you her
renin aldo shortly, which corroborates that
impression. Sorry. While we’re waiting, did
you say you did a renal ultrasound as well? No. So we don’t tend to
go to renal ultrasound, so first line will be
cross-sectional imaging. Okay. You didn’t say
what was electrolytes, potassium. So potassium was
in the normal range. So again, when we look at
her renin and aldosterone and the drugs that
she’s on, the fact that she consistently has a
normal potassium, she’s never at any stage
been hypo- or hyperkalemic. [inaudible] So in terms of
the rest of her workup, as routine in 2012 when
she first came to clinic, we would do
urinary catecholamines. This was done on two
separate occasions and negative both times. I’m sure a lot of us have
moved on to plasma readings now because of
high sensitivities. We haven’t repeated that
because of the two negative urine collections. Her renin-aldosterone
was taken on medication, and I’m sure lots of
you are unsurprised, given the result, she’s on
a high-dose beta blocker, dual ACE and ARB blockade, and this was what
you would expect. Microphone two, yes. Thanks. David Lappin from Galway. I have a couple of comments
on the endocrine workup. I am a nephrologist, but
did you deliberately do a CT adrenals in advance of
doing the endocrine workup, or were they incidental? Did you incidentally look at
the adrenals on the angiogram? When we request our
cross-sectional imaging in all of our patients,
we ask for both. So we asked if we can have a
look at the renal arteries, and also do a specific
scan to look at the adrenal glands as well. As you know,
scanning the adrenals without the biochemistry is
probably putting the cart before the horse. The other thing
about the catecholamines, did you check urine
metanephrines as well, because you’ll miss a small
percentage of PPGLs if you just do catecholamines and
leave out the metanephrines. Yes, we did. Negative. I think the cart before
the horse argument is interesting. I think the issue we
have with a lot of patients, and I’m sure
everyone else does, is what do you do with
somebody that’s already on multiple agents, which
tends to interfere with our interpretation of the
renin and aldosterone, as opposed to newly referred
patients where we can switch them over to alpha blockers
quite easily and then get a more interpretable result. So I’m obviously
interested to hear what your take on that would be.
Yeah. So there was a Dutch
study recently showing that stopping medications in
these very resistant people, in order to do the tests,
doesn’t result in any patient harm. Now, occasionally there’s
the odd patient I think you can’t do it, but in the
majority of patients you can switch. I think the range of
treatment she was on, beta blocker,
diuretic, ACE and an ARB, this is not just
a dirty screen. I’d call it a
filthy screen. We do try and interpret it,
but I just actually think you cannot interpret that
data. If you were to stop one of her medications
in order to get a better interpretation,
which one would you stop? I think the beta
blocker probably. Yeah. That was our
impression as well, because that’s the only one that’s really
suppressing her renin. In a lot of context when
you’re not sure if they’ve got a lump in their adrenal,
whether it’s functional or not, and you’re in this context, we would obviously stop
the beta blocker and if the renin remained
suppressed, then obviously your index of suspicion for primary
aldosterone is increased. There’s another
subtle point on the calcium channel blockade. Dihydropyridine calcium
channel blockers can block your aldosterone level,
so ideally they should be on a non-dihydropyridine calcium
channel blocker I think if you really want to
be clean about it. Yeah, agreed. The ideal situation
would be to do it on no medications. At this point, we would
stop the beta blocker and ideally wouldn’t be on anything. The finest bullet point
is both you’re extremely worried about this lady
and you don’t want to stop, but you think Conn’s
syndrome is the most likely endocrine abnormality you’re
going to come up with here, but just [inaudible] go to a PET etomidates scan and
forget the rest of it. Yeah, we occasionally send
our patients up to Cambridge for exactly that. I think if we would’ve had
a higher index of suspicion in her case, we certainly
would have done that. Okay, let’s move on. The presence of really low
renin in the face of high dose ACE inhibitor certainly indicates
volume suppression. The presence of low aldosterone
at this context suggests something else that
substitutes aldosterone and causing hypervolemia,
[inaudible] tension, like licorice NSAIDs, so I think at this
point it would be, if not earlier, would be
appropriate to go back to the patient and have some
inquiries about her habits. That’s exactly
what we were thinking, but the trick here would be
to stop the beta blocker. So you know what the renin’s doing in that context. Unfortunately,
when we tried to do that, she complained of
severe headaches, and essentially started taking
her beta blocker again. We tried this on two
or three occasions, and were frustrated by the
fact that we couldn’t get at least a look at her renin
without the beta blockade suppressing her renin. So obviously take a comment about the renin-aldosterone
situation. Dr. Kahan Before
discussing whether you can analyze renin-aldosterone
on drugs, it would be interesting to
know whether she has any drug concentrations at all. Otherwise, it’s a
theoretical discussion. On that line,
what about heart rate on your ambulatory blood
pressure measurement? Because she should be
bradycardic if she takes her beta blockers, and if not,
that would help. Finally, we stop all
anti-hypertensive medications and it hasn’t happened
anything since 1987. If you are on five drugs
and 200 systolic, nothing will happen. Yeah, I think on
the last point, we could’ve been braver,
certainly. Some resistance from her. I’ll show you some
data in a second about her heart rate. Yes, we would love to have,
we don’t have the routine facility to measure
by mass-spec patients metabolites the
drugs they’re taking. We know from
studies in Leicester, for example, that 50%
of patients, routinely refer to
this type assenters, do not take their
medications as prescribed. There’s other data that it
could be even higher than that, and I’ll show you
some of that in a second. What we do actually is we
bring people in to our day unit and watch them take their
tablets and document that. So we do
directly-observe therapy. It’s a lower-fi
version of doing that. So I did this personally in
a building that’s actually now been knocked down at UCL. Because of previous experience, we are at this stage, when
we did this in around 2013 with her, we were reluctant
to give her all five of her medications at once. I’ll show you
why in a second. So we gave three of
them at the offset. Now, she’s got a
really high blood pressure initially, and her
heart rate’s 85. Now, after the
three medications, her heart rate did come down in response to the beta
blocker to 70. Her blood pressure’s
not done very much at all. With the second two drugs,
which now include a vasodilator, we do have a
slight improvement in her blood pressure. But five medications in… I watched her take them,
I’m pretty convinced. We left at 3:00 with a
blood pressure of 205/105. So we were pretty
convinced she had true resistant hypertension. Now, the next question
I wanted to throw to the floor, really, is
what would you do? What do you routinely
do faced with this? I personally don’t think
this is that common to find a patient like this who
really does appear to religiously take her tablet. She’s very trustworthy. We saw that her blood
pressure doesn’t drop. We haven’t found
a secondary cause. She’s already on
five medications. This is 2012-2013. What would you
have done with her? We discussed about the
use of beta blockers. Why don’t you use, for instance, non-dihydropyridine
calcium channel blockers? Because when we
consider some biochemical measurements in those patients, we normally replace beta
blockers or some other drugs with non-dihydropyridine
calcium channel blockers, verapamil, a slow-release
verapamil, or the slow-release diltiazem. When you use those drugs
and you may see probably even different blood pressure, or in fact in the meantime
you can also perform biochemical measurements
more accurately. That’s our routine work
when we have a resistant hypertension for some
measurements of those secondary hypertension. Thank you. That’s a good idea.
I’ll take that back. You didn’t mention the calcium. I’m presuming it’s normal. Yeah. Our guideline would say
add Spironolactone. She had an aldo-renin
ratio of over 800:1. Your respect to whether she
was on a beta blocker or not. She’s not responding to a huge
dose of an enalapril anyway. The other thing is either
up the diuretics and/or up the calcium channel blockers. You’re only on 10 of Lercanidpine You could’ve used 20. You could’ve used
nicardipine 45 BD. You could’ve
added torasemide. The guideline would say
add Spironolactone first, up the diuretic and
then up the CCB, and way down on the enalapril, which above 20 probably
didn’t achieve anything. Yeah, and we also… You can’t be 100% sure
she doesn’t have primary aldosteronism,
given the biochemistry, and we actually thought along
the same sort of lines. So we initially
added spironolactone, and also we saw that slight
response to the vasodilator, so we also thought some
additional vasodilatory therapy with an alpha
blocker might be useful. But this was also 2012-2013, and I want to just switch
things up a bit now and suggest something else
that we thought about at that time point. I think the discussion is
now going to move on away from the case. But specifically, and that is, should we consider her for renal
sympathetic denervation? Now, that was because
at this time point, this was very much en vogue, as I’m sure you
can all remember. Just to go back quickly to
remember the underlying idea here, it’s that by
various mechanisms, sympathetic tone both to
and from the kidney help to drive up blood pressure. Then if we can endovasularly, or I’ve heard new technologies, other sort of transcutaneous
technologies now ablate these
arteries that we may affect a reduction in blood pressure. That was the original concept. If you’ll recall, the very
first catheter that was brought onto market was
the Symplicity catheter, and the idea was that
you would ablate in a corkscrew manner
each bilaterally, both renal arteries,
in order to ablate these. Now, I’m just going to
take you through the two studies that came out at
this time point that we had in our hands to help make
this decision whether we should try this
on our patient, this new emerging therapy. I can see a queue of people
at the microphone there. Is this a good place to stop? Do you want to stop here? Please, let’s take
the first question. Before going to renal
denervation, other things I would
consider to optimize their medical treatment,
the patient is taking enalapril once a day. We know that enalapril
doesn’t last 24 hours. The same for losartan. So, first off, all I
would optimize the medical treatment. Even like this, I think
this won’t be enough and you will need more drugs
like spironolactone, doxasosine, anything else. But the first step is optimizing the medical treatment. Agreed. Second, yes? I agree with spironolactone
as the next line, but before even accepting
the data that you had for renal denervation at
that point in time, I mean the trials
were very poor, patients were
their own controls, et cetera, and it wasn’t til
hypertension three that we actually had proper
trial data for it. You still, even with
that poor data, and a lot of people went ahead
and did renal denervation, you still hadn’t ruled
out a Conn syndrome. My own experience with this
is when I’ve taken the renal denervation patients
into my clinic, we’ve diagnosed Conn syndrome
in a lot of those patients. It’s never been clear to
me at any point in those studies that the patients
were properly worked up. Yeah and I think that’s the
beginning of a discussion I that I think I want to
throw a bit wider when we go through the actual
data from the trials. The advice that
was given nationally, which I’ll show you, in the
UK we had some guidelines published, which I’ll touch on. Yes, we’ll discuss some of
those points a bit later. Yes? Hello. Phillip Delmar from Belgium. I’d like to go back
for just one second on the workup of this patient. First, do you have an ID of
her urinary sodium levels? I’m afraid I don’t, no. Because it’s a known fact
that low sodium diet is more effective in resistant
hypertensive patient than in general hypertensive
population, so that could be one point. My second question is, do you
at any point in the care of this patient considered
doing polysomnography to look for sleep apnea syndromes, since you mentioned
that she was a non-dipper, so I think that might
have been interesting. Yeah, so in terms
of the sleep apnea, she doesn’t have the body
habitus of somebody with sleep apnea. I think we don’t
do that routinely. We essentially ask
questions about sleep snoring, then we might use
something like the Epworth, or a first-line questionnaire. You’ll see from our cohort
that we do diagnose sleep apnea in our patients
reasonably commonly actually. It seems to be an
increasingly recognized… I think in retrospect,
she’s a non-dipper so maybe. She just doesn’t have
that body habits or those symptoms really to
suggest sleep apnea. Because you have these
large Brazilian series from a few years ago that
showed in that case, in that series at least,
that something like 60% of patient with resistant
hypertension had sleep apnea syndrome. Yeah, and I think we’ll
find with these patients that have got the
metabolic syndrome, they’re diabetic,
they’re hypertensive, they’re obese, they’ve
got resistant hypertension. I think we’re increasingly
recognizing that if you can get them on a CPAP
machine at home, then actually their
blood pressure comes down. You see that with their nocturnal blood
pressure improving. Yes, please. Rapid remark. About the treatment
of the patient, she had a double blockade
of renal[inaudible] system, which resulted in [inaudible] and I think that in 2013
we had on-target results that showed that there was
an increased mortality, cardiovascular mortality in
patients with double blockade. Why did you not stop there? Yeah, I failed to mention that
the spironolactone replaced. I can’t remember if
it was the ACE or ARB, but we took her off one. As you quietly mentioned,
there was no trial data supporting being on dual ACE and
ARB blockade in hypertension. Just for in department,
you used I think the rapid releasing form. Can you not switch
to a long-acting form, 1 or 5 [inaudible] that has, or chlorthalidone that
is a longer-acting diuretic? Yeah. I think those are some
brilliant suggestions, particularly the idea of
looking at the half-life of these drugs and making sure that she’s getting 24-hour
anti-hypertensive. Definitely take
that back as an idea. Could I also ask you about
some of the more potent classical vasodilators,
like hydralazine. Did this ever come into
consideration in terms of your management? So the discussion was
around vasodilating her. We thought that we’d
start with an alpha blocker. I mean the idea here
is obviously that she’s a good patient,
in the sense that she takes her medication.
She’s eager to participate. So we didn’t
specifically say, “Should we use hydralazine?” We started off with doxazosin. Okay. Yes, please? Then we’ll move on. Thank you. Roberto Pedrinelli
I don’t understand something in this case,
in the overall context. Because this patient had
no hypertrophy, right? She had no albuminuria, right? Presumably, she had no
hemorrhages nor exudates, right? Just Grade 2. Probably. So how do you fit in the
overall picture of this? What do you think about- Well
I think it’s quite remarkable. Like I said,
five years later, having just read her most
recent clinic letters and seeing her repeated echo
and uria and electrolytes, she remains essentially with
very little target organ damage, despite persistent blood
pressure being very elevated. Her most recent blood
pressure is more controlled, so we’re looking at more
160/90 than we were 200/110 with the changes
that have been made. But you’re right, I don’t have
any bright ideas about why. This is why she’s quite
an interesting patient, that she has clearly
sustained hypertension without that end organ damage. We had a case like that,
and we used urapidil, infusion of urapidil,
and we found a very brisk hypertension response
[crosstalk] hypertension. Probably what you did,
to exclude the pseudo-hypertension,
what do you… You gave drug
by mouth, right? Yeah. Was this enough for you? Yes, probably should
be [crosstalk] She had all five of the
prescribed medications that she was referred to us on
at that time point, and her blood
pressure didn’t change. We haven’t repeated that. Her 24-ABPM has improved. We did discuss whether
we should bring her in for intravenous therapy to
demonstrate that we can lower her blood pressure, but we didn’t see a
reason for that. She’s not at any point presented
in an accelerated fashion. Okay, but just… Let’s go on to Jean Paulo. Do you want to make a
comment before we move on to see what happens with
the renal denervation? Yeah, I’m just conscious
of the time and then I’ll obviously have to speed
up various aspects of it. So last comment. Actually, the question
is did you up titrate your spironolactone adequately,
and the reason for that is approximately 1/3 of the
patients do respond to a dose between 25 and 100. Then another 1/3
to 100 and 200 mg. Then 1/3 of the patients
require a higher dose. If you do so, you probably
can withdraw your beta blocker, and that would be
very critical because if you would then find very low
plasma renin activity, or active renin. With an ARB and spironolactone, then you will have very
strong evidence for doing an adrenal vein sampling
during the treatment, which would probably show
you a lateralized secretion of aldosterone. I think, again,
that’s a very good idea. I mean, let’s control
her and go back and do an etomidate scan or
something else, given the ratio’s
very high despite the low aldosterone,
or medications. Very dirty,
filthy, as we found. Let’s see what happens
as you move on with your intervention. Now, what I’m
going to do, I’ll be, bear in mind we don’t
have very long left, I’m going to essentially
do this very quickly. So you all recall that
Symplicity hypertension one had brilliant results. This was a
non-blinded trial. Only had one arm,
which was the treatment arm, and what was interesting,
apart from just demonstrating that patients
dropped their blood pressure very significantly, they
also performed noradrenaline spillover, which are very
difficult technique to perform. But demonstrated that on
average they dropped that by 47%, and that
becomes important later. It was very well tolerated
in terms of safety. Now, this was a first
in man proof of principle. The second trial that came
out the year later was the first randomized
control trial, so now we have a control arm. Again, not blinded. If you have a look
at the entry criteria, our patient would’ve
fit the entry criteria. Unfortunately, she can’t
get in because of her ICD into a lot of these studies. But the blood pressure
that got you into the trials was 160 systolic in the clinic. There was no ABPM
entry in this context. One-to-one randomization,
and they looked at office blood pressure
at six months. They also did a secondary
outcome at 24 hour ABPM. Again, very startling results. The sort of thing when you
look and get a blood pressure reduction of 32 mm
of mercury with our patient, whose blood pressure
is sky-high, you think okay, maybe
this is going to be useful. There was also a drop in ABPM, and again, the complication
rate was small and this was well tolerated. So, in January 2012,
so this was around the time this patient was in our clinic, there was some
guidelines produced by the National Institute of
Clinical Excellence on the basis of two
non-blinded trials. And they obviously recognized
the limited evidence, but suggested this could
be done in routine clinical practice if there was
clinical governance in place, the patient understood
the uncertainty, and there were mechanisms in
place for order and review. It also had to be overseen
by a multidisciplinary team, or MDT, with
hypertension specialists and interventionists. This was supported by
a statement from many societies in the UK, including the British
Hypertension Society, and what they did was
use the same criteria. So if you want to do this,
use the same criteria of the trials, but they fleshed out
the entry or the exclusion criteria in a way that we
thought was quite robust. So it mandated using
ABPM to rule out white coat hypertension. It mandated a
measure of concordance, which was not included
in any of the trials. It mandated exclusion
of secondary causes with suggestions of
first-line investigations. Again, that was not in the
protocol for the two trials. So we thought this sounded
quite potentially useful. We got busy and we got funding
arranged in our hospital. We wrote patient
information leaflets. We had some radiologists
and cardiologists go to Europe to be trained
in how to do this, and we established a regular
MDT to look at patients. I know you’ve been
standing there patiently. I think it’s a good time. Is it a question about… Well, unfortunately is a
question regarding the treatment,
or mainly a comment. The fact that the patient
has low renin can also be a sign that the patient
probably was volume overloaded, so a different
strategy would be to use as many diuretic as possible to assure a right
patient at volume. Another comment is the fact
that there was no target organ damage, it can be
related with the fact that aldosterone itself was
low and aldosterone is one of the major
mediator of fibrosis, so that can be probably
part of the story. So what I would have done
with this patient is to add chlorthalidone,
amiloride, spironolactone, and see what’s happened. I agree. I mean obviously with the beta
blocker caveat in there, if we’d have managed
to get a reading off and she was a low renin
hypertensive patient, which we know makes
about 20% of this cohort, it might change our thinking. I just want to share
with you that this patient description or profile
reminds me of those who responded very well
to renal denervation. Very well. Okay. [crosstalk]
So, on that note. We wanted to know, did we
actually have any patients that we would offer this to,
that would fit the criteria. So we looked prospectively
at our cohorts. These are all the patients
that were booked in to see us already in 2013,
so mainly follow-ups, the occasional new referral. So we had a total of 298
patients that we looked at to try and find some
of these patients. The mean age was 62. They’re relatively
well controlled as a mean. We put them through this
pathway that was derived from the guidelines. So they had to have a clinic
blood pressure more than 160, a 24-hour ambulatory
blood pressure systolic daytime
average of more than 150, and that was again
from the recommendations. Rule out secondary causes. Bring them into our day
unit to see what happens when we give them drugs. So of our 298, about 10%
had a clinic blood pressure more than 160, but the vast,
vast majority of those had either white
coat hypertension, now I’m using the definition
of 150 because that was what was in the guideline, or an
underlying secondary cause. That three were excluded
for various reasons, which left us with
only four patients to bring into the day unit. Of those, two failed,
which means they took their medications and their
blood pressure dropped, as you would hope it would,
which means they don’t take their medications. So we were left with
two potential candidates, and the lady I presented
to you was one of those. Just a little
bit more detail, the most common secondary
cause in this cohort was renal artery stenosis. We had some Conn syndromes,
some reasonably severe CKD, and obstructive sleep apnea. One patient had
coarctation of the aorta. Really interestingly is
over half of the patients that we did ABPM in dropped
their blood pressure to less than 150 systolic
in the daytime, so they were all excluded. So this is what happened
with one of the patients that came in. Took four medications at 9:00, and by lunchtime had a
blood pressure of 130/74, so a very nice response. I mentioned we
changed our protocol. We learned the hard way, as I’m sure various of you have. We gave, again,
four medications to another patient who initially started
with a blood pressure of 160/90, and unfortunately
at lunchtime, we were fluid resuscitating
the patient who became syncopal and blood pressure
had dropped significantly. So at this point, we had to administer
intravenous fluids. I don’t know if any of
you have had a similar experience when you
bring patients in. It’s quite scary. We learned our lesson. We don’t do this anymore. Fortunately, by the end of play, blood pressure was up and they were no longer symptomatic. So we had these two
candidates in October 2013. You’ll recall
what happened next. In January 2014, Medtronic,
who make the Symplicity catheter announced
that their next trial, which was the first
blinded sham control trial, failed to meet its
primary end point. What do you do? Immediately in January,
the joint consensus groups came out with a statement
recommending a temporary moratorium on renal
sympathetic denervation until we know a bit more,
so we had to stop our very new, yet to enroll a single
patient service for renal sympathetic
denervation. In fact, after
review of the evidence, that temporary moratorium
has become permanent. We are still not allowed
to perform renal sympathetic denervation outside the
context of a clinical trial. So just a brief look
at the trial that caused a lot controversy. The trial was
performed in America. Now, up to this point they
were all performed in Europe or Australia, and they have
the same inclusion criteria, but they included
the sham procedure, which was really important. They did not have
an analysis based on noradrenaline spillover
in this trial. The results, initially,
if you look at the group that had renal
sympathetic denervation, they did drop their
blood pressure by 14 mm of mercury. Not as impressive as
Symplicity Hypertension 2, but that wasn’t the issue. The issue was the sham
group also dropped their blood pressure to the
same degree and there was no statistical difference. Neither was there a
difference in ABPM. Complication rate was not
statistically different. So, a question that has
taxed us as a community was why did this fail, and we’ve heard a few
potential reasons for that. Does anyone have any other
thoughts about why that trial specifically
failed to demonstrate any benefit of renal
sympathetic denervation? A couple of points. I think the average
number of interventions per interventionist was two, so probably the wrong
people were doing it. So no experience from the
people actually performing it. The other thing, I mean there is questionable
science around it still. That hasn’t all been tidied up. The other, the latest
thought on it is that where they were denervating, they weren’t deep enough
into the artery. They’re saying you got to
go deeper into the artery. A lot of people probably
weren’t done completely in a spiral manner. Again, clearly the workup
may have been better for these patients. So technical problems
with the procedure, it not actually achieving
denervation as advertised. Jan. I agree, no experience. But where I don’t agree
is I don’t think you need a sham procedure because
you can measure the blood pressure by advice. You can do 24-hour
ambulatory monitoring. You can have a
central reading station. The person or the device
who reads the blood pressure doesn’t know whether
it’s an intervention or whether it’s a sham. So I really don’t think that a
sham procedure is needed. You’re not alone. There’s a recent paper that
was a meta-analysis that looked to that question
as whether the sham design is necessary. I think from the
FDA point of view, they’ve mandated, obviously,
a sham control trial because… There’s been a lot of
discussion in the literature, and you can go back, and a lot of the things
that have just been said have been brought up. One of the questions is
does it actually work. Is this approach
valid scientifically? Or that this trial should be
lauded because it showed the previous flaws in the other
two trials by not having a sham that led to a
positive result. There’s questions about
the inclusion criteria not being rigorous enough so that we
are getting a very dirty cohort, and that the technical success
of the procedures we heard was impossible to measure
because there was no outcome. So the arguments are that
there is the concept of the Hawthorne effect. So patients in a trial who
know they’ve had a novel new therapy are going to start
taking their medications, and that there is a reverse
effect where those that didn’t have the new intervention stop
taking their medications. That’s been proposed. Why are these patients
so susceptible to the Hawthorne effect? It’s because they
don’t take their tablets. So the sympathy trial,
which measured people’s urinary and plasma drug levels
without their knowledge, showed that up to 80% of
patients were either completely non-concordant
or poorly concordant, which means there’s lots
of scope for medications, the amount of medications
that are being taken to change during the
course of the trial. Importantly, none of the
sympathy trials included a measure of concordance
at any point. There’s also this question
of stability of medications. If you’re only on a regime
for two weeks before you enter a trial, are
you really the sort of patient we want to be entering,
who’s really difficult to control is on a stable regime? 40% of patients in the
trial had their medications changed at some point, so it
really does throw into sharp relief that particular problem. Now, when we applied
the guidelines, which was directly
observed therapy, ABPM, rule out secondary causes, we could only get 6% of our
cohort with a blood pressure of 160 or more in
our new service, which never really started. In the trial, they were
getting near to 40%. So I don’t think they
were really working up these patients sufficiently, and they got a really
poorly defined cohort. Even with that problem,
we heard they needed 88 sites, so there’s only six procedures
done in each site. If we look at
technical success, in Symplicity
Hypertension 1, they achieved a 47% reduction
in noradrenaline spillover. But in animal models,
you can get near to 100%. In another small study
in patients that had renal sympathetic denervation,
the spread of effect was enormous. So I think it’s become
quite apparent that the original method of
denervating the kidneys in these trials was not
effective at actually ablating the renal sympathetic
nerves appropriately. So it’s more complex. It’s much more difficult
than we originally thought, and not only is it
more complicated, the interventionists in the
third trial did not adhere to the protocol. The protocol was 4-6
ablations in a corkscrew manner in both
renal arteries. Only 5% actually had a per
protocol ablation performed. Again, that was
because of inexperience. 34% of interventionists
performed just one procedure. How can they possibly be
expected to get it right the first time? So there’s lots of issues. Now, some of the lessons
that we propose is should we have a sham, but people
have questioned that, that we need to include when
we’re designing these trials. A robust measure
of concordance, have a much longer period
of stability of medication. Maybe uniform, make a
uniform regime that people should follow. We probably should only
use centers of excellence. Maybe we should not have
been trying to get involved in this story at all. We should’ve been sending
them down the road to where they do them and have done
them as part of the trial. Part of our
naivete was that we, and our interventional
radiologist is that there’s something new and
fun that we can do. But actually the evidence
suggests that probably you should send it to experts. Some of these ideas have
actually been tested in more recent trials. So there’s been two
more sham control trials. These were
performed by experts, people that had
done 20-30 procedures. In the research trial,
every single renal sympathetic denervation
was performed by the same interventionist, yet they
still failed to demonstrate a result. But they didn’t
measure concordance. I’ve already mentioned
the sympathy trial. We know that concordance
is a real problem. When they analyze the data
according to patients that had a stable regime
versus those that didn’t, they started to
see a signal, but only in the patients
that had stability of their regime throughout. So that shows you how, when
you have such a cohort that are not taking
their medications, it’s impossible to see
the signal from the noise. In terms of technique,
in a porcine model, as we’ve just heard,
where you ablate is really important, so the nerves
become much more accessible distally, so if you ablate
the main artery but also the branch arteries,
you can then bring down noradrenaline spillover to
the sort of levels that you would think,
physiologically, are going to be relevant
to a clinical outcome. So, before I go, this is
my penultimate slide of two minutes, I think. Yes, if we could
just have comments, brief comments,
so we can finalize. Of course, all the points
that you made are important and extremely relevant,
but the major fallacy is that resistant hypertension is
not a physiologic entity. It is a mixture. Some patients are
resistant to my treatment. That does not necessarily
mean they will be resistant to your treatment,
and vice versa. So I think that major
fallacy is the idea that you can take a bag of all sort
of things and treat them with a specific [crosstalk]
Single treatment, yeah. Moreover, I would like
to cite a study from [inaudible] I don’t know. quarter of a century
ago or more. Who worked up,
I don’t know, 7,000 people for
secondary hypertension, and ended up finding all
sort of findings and the majority were not free from
anti-hypertensive treatment regardless of the therapy. So longstanding hypertension is quite an elusive challenge. Okay, we’ve got one
and a half minutes left, so brief comment and
then we can wrap up. Thank you for all
these clinical trials. But there is one trial that was
not open label trial. But with a
stepped treatment, doing four weeks-
The French trial? Yeah, the denier HTN.
Yeah. But in which there was a step
treatment for four weeks before confirmation of
resistant hypertension, and then they were
randomized to a intervention or to a stepped
escalation in treatment. They were also a [inaudible]
dosage of the treatment and in this
trial it was positive, and there is a significant
difference between the two groups, and this difference
was more important in those who were during
the treatment. So even if there is
no sham procedure, we can talk
about this result. This is really important. I think the key
message of that trial, and it’s been incorporated
into the latest Spiral trials, which is medication
stability is vital in order to see a signal. What I wanted to show is,
because we’re all missing right now the
update from Spiral. You’ve all
chosen to come here. You didn’t want to go to renal
sympathetic denervation, but you got some anyway. Apologies for that. But what they’ve tried to
do is integrate all of the lessons that have been
discussed in the literature, including what
you’ve mentioned. The first is their new catheter. It’s supposed to be
easier to achieve, instead of a 90% failure rate, they should get 100% success, and that it would be applied
to the main and branch arteries. They’re including lots of things
that we all think are important. So patients are going to be
on three specific medication classes, in the same way it
was done in the French study. There’s going to be six weeks
stability instead of two. They are including
directly observed therapy. They’re also measuring drug
levels throughout in urine and plasma by mass spec. They’re restricting to
one procedure as per site, so that people get the
experience that’s required. There is an outcome
at three months, but they’re also going to look
at an outcome at 36 months. So they really have
integrated all these lessons. Therefore, this trial,
you would hope, should be rather definitive. Although, I’ve just spoken to, well I’ve spoken to
Medtronic reps and apparently this
may come out at ESC, which would be interesting. But interesting to have
an off-med study as well. So how would you deal with
this problem of concordance? Well, maybe you should
perform this on people that aren’t taking any
medications at all. I think this is
quite controversial, but this is what
they’re doing. They’re taking patients
that have the same blood pressure off medication
and performing the study. The rationale for this is
it’s the cleanest cohorts where, if you’re going
to see a difference, you’ll see a difference. This is not unique. Other companies who
have slightly different technology are
doing similar things. In fact, if you are interested, [Recor] have their
update tomorrow about their radiance trials. They are using an on and
off-med strategy as well, really I think to try and offset this idea of
poor concordance. So from our patient’s
point of view, high blood pressure
is slightly better. I’m going to take back
some of the excellent recommendations from
everyone on the floor about how we might
manage her better. She can’t be enrolled in these
studies because of her ICD. I think as a center,
we feel that we were very naïve in the first instance of
getting involved in this bandwagon, because we
didn’t wait for a blinded randomized control trial
before setting this up, and I think that’s
a real lesson. If this had been a new drug, there is no way that it
would’ve gotten this far and people have got so
involved like ourselves. So we’ve learned our lesson, and I think the lessons
are generic. I think these need to be
applied to any new device or approach to hypertension
moving forward. So, obviously I’m out of time. Thank you very much indeed. Dr. George,
thank you so much. With that, we will
close the session. Thank you.

Be the first to comment

Leave a Reply

Your email address will not be published.


*