ALLHAT: Optimal firststep therapy for hypertension

1
ALLHAT Optimal first-step therapy for
hypertension

• Eric J Topol MD Provost and Chief Academic
  Officer Chairman, Department of Cardiovascular
  Medicine The Cleveland Clinic Foundation Clevela
  nd, OH
• Robert M Califf MD Professor of
  Medicine Associate Vice Chancellor for Clinical
  Research Director, Duke Clinical Research
  Institute Duke University Medical
  Center Durham, NC

2
Randomized designof ALLHAT
Amlodipine Chlorthalidone Doxazosin Lisinopril
High-risk hypertensive patients
Consent / Randomize (42 418)
Eligible for lipid-lowering
Not eligible for lipid-lowering
Consent / Randomize (10 355)
Pravastatin Usual care
Follow for CHD and other outcomes until death or
end of study (up to 8 yrs).
3
A horse race

• Each class of antihypertensive is represented by
  a drug, and the losers drop out as events are
  accrued
• Primary end point fatal CHD or nonfatal MI
• All major clinical end points were measured in
  minimal detail

Califf
4
Secondary drug protocol
5
ALLHAT Trial design

• 42 418 patients age gt55 with hypertension and 1
  additional risk factor
• 623 sites
• United States
• Canada
• Puerto Rico
• US Virgin Islands

6
Doxazosin CVD end point
doxazosin
chlorthalidone
Cumulative event rate
12,990 7,382
9,443 5,285
4,827 2,654
2,010 1,083
Years of Follow-up
C 15,268 D 9,067
JAMA. 20002831967-1975
7
ALLHAT Primary end point
JAMA 2002 2882981-2997
8
Lisinopril secondary end points
JAMA 2002 2882981-2997
9
Amlodipine secondary end points
JAMA 2002 2882981-2997
10
ALLHAT Fasting glucose levels
JAMA 2002 2882981-2997
11
Stroke risk Lisinopril vs chlorthalidone
JAMA 2002 2882981-2997
12
ALLHAT Glomerular filtration rate
JAMA 2002 2882981-2997
13
ALLHAT-LLT Primary results
JAMA 2002 2882998-3007
14
ALLHAT-LLT Disappointing

• Second largest statin trial after HPS
• "It's disappointing that it didn't provide
  true consistency and only with this
  bouillabaisse pooling stuff do you get the same
  relative effect."

Topol
15
ALLHAT Points of contention

• Why did lisinopril increase heart failure and
  stroke?
• This is directly opposite of the results from
  HOPE
• PEACE and EUROPA are looking at ACE inhibitors as
  a key preventive tactic
• "This backfired terribly in ALLHAT."

Topol
16
Active control trial

• The other drugs were only less effective than the
  diuretic, not increasing risks for the patients
• The "soft underbelly" of HOPE was whether the
  patients were being adequately treated with
  regard to their other risk factors
• If EUROPA and PEACE are negative, either HOPE was
  wrong or ramipril is "a magic potion"

Califf
17
Using less ramipril

• I've gone from requiring ramipril use to making
  it optional
• "I think we have to say this is a piece of data
  that moves back toward less radical enthusiasm
  about the ACE-inhibitor class."

Califf
18
ALLHAT Blood pressure
JAMA 2002 2882981-2997
19
Genetics of hypertension

• Studies suggest the genetic defect of essential
  hypertension alpha adducin Gly460Trp would be
  particularly responsive to thiazide diuretic

Topol
20
Stroke risk Lisinopril vs chlorthalidone
JAMA 2002 2882981-2997
21
ALLHAT Blood pressure
JAMA 2002 2882981-2997
22
Amlodipine secondary end points
JAMA 2002 2882981-2997
23
Edema or heart failure?

• There was no objective measure of function to
  diagnose heart failure
• A substudy was commissioned to have records
  independently reviewed
• All the results are not in, but so far the
  substudy suggests that there is more than just
  edema going on

Califf
24
Surprising increase in heart failure

• The increase in heart failure for both classes of
  drugs (ACE inhibitor and CCB) was a very
  surprising finding
• Lisinopril 19 increased risk
• Amlodipine 38 increased risk
• "You would have thought both drugs would not
  have done this."

Topol
25
Bad choice of second drug

• Critics say the second drug after ACE inhibitor
  would be a diuretic, forbidden by the trial
• Most doctors in the US probably don't use a
  diuretic as the second drug
• "I think no matter how you slice the loaf here
  the answer is that the underused diuretics,
  which are a lot cheaper, are at least as good
  and almost certainly better."

Califf
26
A class effect?

• Most doctors use hydrochlorothiazide as a
  diuretic
• This could be a chlorthalidone-specific result,
  you can't be sure
• "We have examples where drugs in the same class
  don't get the same results."

Califf
27
Striking secondary outcomes

• The media loved that a cheaper drug came out
  better
• I wasn't enthusiastic about the trial when I was
  on the NIH advisory committee reviewing the
  trial
• "The secondary outcomes made for all the spice
  here. If you were to just go by the primary
  outcome, though, you wouldn't be able to
  differentiate the treatments."

Topol
28
Public health

• Why not use a cheaper drug that is just as good?
  It's a dominant treatment
• A company trying to get labelling with this
  primary outcome might have trouble getting
  approval from the FDA
• "We have examples where drugs in the same class
  don't get the same results."

Califf
29
An easy choice

• There is a value judgment being made among the
  secondary outcomes
• "I think the majority of people, if you said,
  'Look, I can give you this thing for 2 cents a
  day, or I can give you this thing for a buck
  and a half a day, and here are the expected
  outcomes, which would you buy?' I don't think
  that's a hard choice."

Califf
30
Inadequate treatment

• "None of these drugs are very good, they all
  have some untoward effects, unfortunately.
  You're picking your poison in some respects."
• This study reinforces that there is inadequate
  treatment of blood pressure.
• "A lot of people are walking around with very
  high blood pressure still, despite therapy."

Topol
31
Lowering blood pressure

• "I think that people that are most critical of
  doctors trying to lower blood pressure are
  people that have never actually worked in a
  clinic trying to get blood pressure down."
• It takes the doctor and patient working
  together to get blood pressure down
• We usually need more than 2 drugs

Califf
32
Genomics approach to hypertension

• Genomics will allow us to move past the trial and
  error approach
• "37 billion a year it costs to treat
  hypertension and we're not even doing a very
  good job of doing it. We've got to have a better
  strategy and almost any strategy would be
  better than what we have today."

Topol
33
Interpersonal approach

• People will round numbers down for patients who
  are frustrated at not getting hypertension under
  control to avoid adding more drugs drugs
• Lowering blood pressure is very complicated and
  interpersonal

Califf
34
Two thumbs up

• Topol "Don't you think this is as good as it
  gets for hypertension and clinical trials?"
• Two thumbs up
• Califf "I think it's as good as it gets."
• Two thumbs up

35
A new approach

• "The system we have now, where companies not
  only fund trials but decide what the questions
  are is not the right way to do it"
• Doctors and patients want to know which is the
  best choice among the treatments that work?
• Most companies avoid head-to-head trials and try
  to game them even when they agree

Califf
36
Honest broker

• NIH or other agencies as an honest broker is the
  model to pursue in the future
• "If a drug is a winner it ought to prevail in
  a direct comparison without the type of
  engineering that can occur with interested
  sponsors."

Topol
37
Pricing

• The price of a drug should be a function of how
  much it contributes
• "The way it is now, people are having to decide
  what to take and what to buy without any
  knowledge in many fields of which one is really
  better."
• In multiple sclerosis, for example, there are 4
  drugs and no one knows which is really better

Califf
38
ALLHAT Optimal first-step therapy for
hypertension

• Eric J Topol MD Provost and Chief Academic
  Officer Chairman, Department of Cardiovascular
  Medicine The Cleveland Clinic Foundation Clevela
  nd, OH
• Robert M Califf MD Professor of
  Medicine Associate Vice Chancellor for Clinical
  Research Director, Duke Clinical Research
  Institute Duke University Medical
  Center Durham, NC