Disclosures I have nothing to disclose Heart Failure for the - - PDF document

1

“Heart Failure for the Hospitalist”

Ronald Witteles, M.D. Stanford University School of Medicine October 22, 2016

Disclosures

I have nothing to disclose

Goals of This Talk

• Focus on real-life clinical scenarios you will encounter
• When to place an ICD?
• When to initiate/stop heart failure drugs?

A Common Scenario

• 68 y.o. man with longstanding nonischemic

cardiomyopathy

• At diagnosis:
• LVEF 20%, normal cath
• EKG: Narrow QRS, no significant abnormalities
• Well-controlled DM II, otherwise healthy
• Now (3 years post-diagnosis):
• On stable comprehensive heart failure medication regimen
• Admitted for volume overload due to excess Na intake
• You have diuresed him & he feels better
• LVEF 30% (stable for last 2 years)
• He’s getting ready for discharge, and your Cardiology

consultant wants to place a primary prevention ICD

2

What do you say?

1) “Makes sense. He has a LVEF <35%, a life expectancy >1 year, and is on a comprehensive medication regimen.” 2) “I could take it or leave it. The evidence is equivocal.” 3) “No thanks. Primary prevention ICDs only make sense for ischemic cardiomyopathies.” 4) “No thanks. Primary prevention ICDs only make sense for LVEF ≤ 25%.” 5) “Of course! I blindly listen to the consultants regardless of the evidence!”

What do you say?

1) “Makes sense. He has a LVEF <35%, a life expectancy >1 year, and is on a comprehensive medication regimen.” 2) “I could take it or leave it. The evidence is equivocal.” 3) “No thanks. Primary prevention ICDs only make sense for ischemic cardiomyopathies.” 4) “No thanks. Primary prevention ICDs only make sense for LVEF ≤ 25%.” 5) “Of course! I blindly listen to the consultants regardless of the evidence!”

What do you say?

1) “Makes sense. He has a LVEF <35%, a life expectancy >1 year, and is on a comprehensive medication regimen.” 2) “I could take it or leave it. The evidence is equivocal.” 3) “No thanks. Primary prevention ICDs only make sense for ischemic cardiomyopathies.” 4) “No thanks. Primary prevention ICDs only make sense for LVEF ≤ 25%.” 5) “Of course! I blindly listen to the consultants regardless of the evidence!”

x

What do you say?

1) “Makes sense. He has a LVEF <35%, a life expectancy >1 year, and is on a comprehensive medication regimen.” 2) “I could take it or leave it. The evidence is equivocal.” 3) “No thanks. Primary prevention ICDs only make sense for ischemic cardiomyopathies.” 4) “No thanks. Primary prevention ICDs only make sense for LVEF ≤ 25%.” 5) “Of course! I blindly listen to the consultants regardless of the evidence!”

3

A Similar Scenario

• 68 y.o. man admitted with an acute MI
• At time of admission:
• Anterior ST elevations
• Emergent coronary angiogram: 100% mid-LAD occlusion
• Undergoes successful PCI with placement of a DES
• Has well-controlled DM II, otherwise healthy
• Echocardiogram 3 days after STEMI:
• Anterior/apical wall-motion abnormalities
• LVEF = 30%
• On beta-blocker, ACE-inhibitor, and aldosterone antagonist
• He’s getting ready for discharge, and your Cardiology

consultant wants to place a primary prevention ICD

• What do you say?

What do you say?

1) “Makes sense. He has an LVEF <35%, an ischemic cardiomyopathy, a life expectancy >1 year, and is on a comprehensive medication regimen. There’s good trial evidence recommending this practice.” 2) “I could take it or leave it. There’s really no evidence either way.” 3) “No thanks. There’s good trial evidence recommending against this practice.” 4) “Of course! I blindly listen to the consultants regardless of the evidence!”

What do you say?

1) “Makes sense. He has an LVEF <35%, an ischemic cardiomyopathy, a life expectancy >1 year, and is on a comprehensive medication regimen. There’s good trial evidence recommending this practice.” 2) “I could take it or leave it. There’s really no evidence either way.” 3) “No thanks. There’s good trial evidence recommending against this practice.” 4) “Of course! I blindly listen to the consultants regardless of the evidence!”

One Final Question…

• The same patient just discussed & the same

scenario…

• You tell the Cardiologist that your patient is not

going to get an ICD as there good trial evidence against it.

• The Cardiologist suggests placing a Life Vest for

the next 3-6 months instead.

4

What do you say?

1) “Makes sense. The evidence shows that Life Vests save lives in this scenario.” 2) “I could take it or leave it. The evidence is equivocal.” 3) “No thanks. There’s good trial evidence recommending against this practice.” 4) “Why would my patient need a Life Vest anyway? He doesn’t even like boats!”

What do you say?

1) “Makes sense. The evidence shows that Life Vests save lives in this scenario.” 2) “I could take it or leave it. The evidence is equivocal.” 3) “No thanks. There’s good trial evidence recommending against this practice.” 4) “Why would my patient need a Life Vest anyway? He doesn’t even like boats!”

Let’s Examine the Landmark Primary Prevention ICD Trials MADIT Trial (1996)

• 196 patients with following characteristics:
• Prior MI
• History of NSVT
• LVEF < 35%
• Inducible VT on EP study
• Randomized to ICD vs. conventional therapy
• Mean f/u: 27 months
• Primary endpoint: All-cause mortality

Adapted from Moss et al. NEJM. 1996;335:1933-40.

5

MADIT Trial: Survival

Adapted from Moss et al. NEJM. 1996;335:1933-40.

P=0.009

CABG-Patch Trial (1997)

• 900 patients undergoing elective CABG
• Age <80 years
• LVEF ≤ 35% (preoperatively)
• Randomized to ICD vs. no ICD
• Mean f/u: 32 months
• Primary endpoint: All-cause mortality

CABG-Patch Trial: Survival

Adapted from Bigger et al. NEJM. 1997;337:1569-75.

MUSTT Trial (1999)

• 704 patients with following characteristics:
• Coronary artery disease
• History of NSVT within last 6 months
• LVEF ≤ 40%
• Inducible VT on EP study
• Randomized to EPS-guided antiarrhythmic

therapy vs. no antiarrhythmic therapy

• Patients in EPS-guided therapy arm could get ICD

if could not find successful antiarrhythmic agent

• 46% got an ICD on this basis

6

MUSTT Trial: All-Cause Mortality

Adapted from Buxton et al. NEJM. 1999;341:1882-90.

MADIT-II Trial (2002)

• 1232 patients on stable medical

therapy with following characteristics:

• Prior MI
• LVEF < 30%
• Randomized to ICD vs.

conventional therapy

• Mean f/u: 20 months
• Primary endpoint: All-cause

mortality

Adapted from Moss et al. NEJM. 2002;346:877-83.

MADIT-II: Survival

Adapted from Moss et al. NEJM. 2002;346:877-83.

P=0.016

DEFINITE Trial (2004)

• 458 patients (!)
• Nonischemic cardiomyopathy
• LVEF ≤ 35%
• NSVT on Holter or telemetry
• Randomized to ICD vs. standard

therapy

• Mean f/u: 29 months
• Primary endpoint: All-cause

mortality

Adapted from Kadish et al. NEJM. 2004;350:2151-8.

7

Results: Not so DEFINITE

Adapted from Kadish et al. NEJM. 2004;350:2151-8.

Results: Not so DEFINITE

Adapted from Kadish et al. NEJM. 2004;350:2151-8.

DINAMIT Trial (2004)

• 675 patients
• LVEF ≤ 35%
• 6-40 days after MI
• Randomized to ICD vs. placebo
• Mean f/u: 30 months
• Primary endpoint: All-cause mortality

Not So DINAMIT: Mortality

Adapted from Hohnloser et al. NEJM. 2004;351:2481-8.

8

DINAMIT: Fewer Arrhythmic Deaths, But…

Adapted from Hohnloser et al. NEJM. 2004;351:2481-8. Arrhythmic Deaths Nonarrhythmic Deaths

Brief Interruption – A Quick Aside About Trial Names & Karma…

BEST Trial: Bucindolol for Heart Failure

Adapted from NEJM. 2001;344:1659-67. * No statistically significant difference

They’ll Never Learn: Another BEST Trial (!) Death, MI, Stroke, or Revascularization in Multivessel CAD

Adapted from Park et al. N Engl J Med. 2015;372:1204-1212.

*Funded by Abbott Vascular, Producer of Everolimus Stents

9

They’ll Never Learn: Another BEST Trial (!) Death, MI, Stroke, or Revascularization in Multivessel CAD

Adapted from Park et al. N Engl J Med. 2015;372:1204-1212.

*Funded by Abbott Vascular, Producer of Everolimus Stents

“What is the optimal antiplatelet and anticoagulant therapy in patients with

• ral anticoagulation and coronary

stenting?”

Learning From One’s Mistakes…

“What is the Optimal antiplatelet and anticoagulant theRapy in patients wiTH

• ral anticoagulation and coronarY

stenting?”

Learning From One’s Mistakes…

“What is the Optimal aNtiplatelet anD anticoagulant therapy in patiEnts with

• ral anticoagulation and coRonary

stenting?”

Learning From One’s Mistakes…

10

“What is the Optimal antiplatElet and anticoagulant therapy in patients with

• ral anticoagulation and coronary

StenTing?”

Learning From One’s Mistakes…

“WOEST” Trial Name Ever… But the BEST Karma!

Adapted from Dewilde et al. Lancet. 2013;381:1107-1115.

Back to ICDs… IRIS Trial (2009)

• 898 patients
• LVEF ≤ 40%
• 5-31 days after MI
• HR ≥ 90 bpm at least 48h after the MI (marker

for high-risk) or NSVT ≥ 150 bpm on Holter moniotring (or both)

• Randomized to ICD vs. placebo
• Mean f/u: 37 months
• Primary endpoint: All-cause mortality

11

Still Not So DINAMIT: Mortality (IRIS)

Adapted from Steinbeck et al. NEJM. 2009;361:1427-36.

IRIS: Fewer Arrhythmic Deaths, But…

Sudden Cardiac Death Other Deaths Adapted from Steinbeck et al. NEJM. 2009;361:1427-36.

SCD-HeFT Trial (2005)

• 2521 patients
• NYHA Class II (70%) or Class III (30%)
• LVEF ≤ 35%
• Ischemic (52%) or nonischemic (48%)!
• Randomized to ICD vs. amiodarone vs. placebo
• Median f/u: 45.5 months
• Primary endpoint: All-cause mortality

SCD-HeFT: Mortality

Adapted from Bardy et al. NEJM. 2005;352:225-37.

12

SCD-HeFT: Mortality in Ischemic CM

Adapted from Bardy et al. NEJM. 2005;352:225-37.

SCD-HeFT: Mortality in Nonischemic CM

Adapted from Bardy et al. NEJM. 2005;352:225-37.

DANISH Trial (2016)

• 556 patients
• Almost all NYHA II (54%) or III (45%)
• LVEF ≤ 35%
• Nonischemic
• Randomized to ICD vs. no ICD
• 58% in both arms received CRT
• Median f/u: 68 months
• Primary endpoint: All-cause mortality

DANISH Trial: All-Cause Mortality

Adapted from Kober et al. NEJM. 2016. Epub ahead of print.

13

Benefit in Younger, Less Sick Patients?

Adapted from Kober et al. NEJM. 2016. Epub ahead of print.

So Where Does This Leave Us?

• Ischemic cardiomyopathy
• If no recent revascularization & stable regimen & LVEF

≤ 35%...

• Place an ICD! (MADIT, MADIT II, SCD-HeFT)
• If recent revascularization and/or acute MI setting…
• Do not place an ICD! (CABG-PATCH, DINAMIT,

IRIS)

• Nonischemic cardiomyopathy
• If stable regimen & LVEF ≤ 35%...
• Unclear what to do (DEFINITE, SCD-HeFT,

DANISH)

• In all groups:
• Must have reasonable life-expectancy!
• Guidelines: At least 1 year (probably too little)

Life Vests What is a Life Vest?

• The only available “wearable cardioverter-defibrillator”
• The setup…
• Patient wears the vest – which is able to detect a

basic electrocardiogram

• Parameters are set to detect VT/VF – based on rate

& morphology different than baseline QRS

• If detects arrhythmia à vibration/verbal warning

that shock will be delivered

• Patient has a chance to press buttons to abort the shock
• Otherwise à gel extruded from electrodes à series of up to

5 shocks, up to 150 J

14

The Potential Issues...

• Bulky, somewhat uncomfortable
• Anxiety!
• No pacing, no anti-tachycardia pacing
• Cost (>$3000/month)
• Only works when you’re wearing it
• How good is the evidence?

The Data

• WEARIT/BIROAD trials
• WEARIT: Ambulatory NYHA III/IV patients, EF <30%
• BIROAD: Patients who might be getting a future ICD

implanted

• Recent MI with VT within 48 hours of infarct
• Recent MI with EF <30% ≥ 3 days post-infarct
• Syncope or cardiac arrest ≥ 2 days post-infarct but

unable to receive ICD

• ICD candidates who logistically couldn’t get ICD for

at least 4 months

• Met ICD criteria but refused ICD implantation
• FDA: Decided to combine both trials into 1 for analysis
• Study endpoint (!):
• Successful resuscitations would occur in at least 25% of

arrhythmic events

Adapted from Kutyifa et al. Circulation. 2015;132:1613 and Feldman et al. Pacing Clin Electrophysiol. 2004;27:4.

The Data

• Results:
• 1/177 patients in WEARIT with successful shock (x2 total)
• 3/112 patients in BIROAD with successful shock (x4 total)
• 2 patients à unsuccessful shocks
• Both had defibrillating pads reversed (not against skin)
• 1 successfully resuscitated by external ICD shock by

EMS

• 23% of patients withdrew from study
• Mainly due to bulk/discomfort of device and/or skin

irritation

• After median of 3.1 months, 12/289 patients had died (4%)
• 6 non-sudden death
• 5 sudden death (not wearing device) – ? Cause
• 1 arrhythmic sudden death from wrong setup

Adapted from Kutyifa et al. Circulation. 2015;132:1613 and Feldman et al. Pacing Clin Electrophysiol. 2004;27:4.

The Data

• Follow-up WEARIT-II registry study
• 2000 patients wore for median of 3 months
• 120 episodes recorded of sustained VT/VF
• 90 aborted by patient à most spontaneously

stopped

• 30 (in 22 patients) à successfully

terminated arrhythmia

• “Only” 10 patients received inappropriate shock

Adapted from Kutyifa et al. Circulation. 2015;132:1613 and Feldman et al. Pacing Clin Electrophysiol. 2004;27:4.

15

Life Vests: My Take

• They probably have a role – but a limited one
• Don’t think of as a surrogate for a permanent ICD
• May make sense in:
• Particularly high short-term risk but possible lack of

need for long-term ICD

• High-risk patient who cannot have permanent ICD

for fixed period

• Probably doesn’t make sense for:
• Routine use in LVEF < 35% while awaiting

medical therapy uptitration

• Routine post-MI/low EF setting

Part 2: Medications, Medications… Medications, Medications…

• 67 y.o. woman with nonischemic heart failure, LVEF 37%
• PMH: Hypertension, gout
• Admitted with volume overload
• You diurese & want to send her home on the best regimen

possible

• Exam at discharge:
• BP 118/70, HR 64, SaO2 98% RA
• JVP 6 cm, no significant extra volume
• I am going to propose some changes to her regimen.
• Your job:
• Would you do this change or not?
• Remember: Polypharmacy – only put pills on if they

matter!

Baseline Medications:

• Metoprolol 50 mg bid
• Enalapril

5 mg bid

• Furosemide 40 mg bid
• Potassium

20 mEq daily

• Rosuvastatin 10 mg daily

16

Option 1: Change Metoprolol to Carvedilol

1) Yes 2) No

Option 2: Change enalapril to sacubitril/valsartan

1) Yes 2) No

Option 3: Add digoxin

1) Yes 2) No

Option 4: Stop rosuvastatin

1) Yes 2) No

17

Option 5: Substitute spironolactone for KCl

1) Yes 2) No

My Answers… My Answers…

1) Yes

My Answers…

1) Yes 2) Yes

18

My Answers…

1) Yes 2) Yes 3) Yes

My Answers…

1) Yes 2) Yes 3) Yes 4) Yes, and…

My Answers…

1) Yes 2) Yes 3) Yes 4) Yes, and… 5) Yes!

Choice of Beta-Blocker

19

COMET Trial

• 3029 patients
• LVEF < 35%
• NYHA Class II-IV
• HF Admission within last 2 years
• Randomized to:
• Carvedilol 3.125 mg bid à 25 mg bid vs.
• Metoprolol tartrate 5 mg bid à 50 mg bid
• Mean follow-up: 4.8 years
• Primary endpoints:
• 1) All-cause mortality
• 2) All-cause mortality or all-cause admission

COMET: Mortality

Adapted from Poole-Wilson et al. Lancet. 2003;362:7-13. P=0.0017

COMET: Heart Rates

Adapted from Poole-Wilson et al. Lancet. 2003;362:7-13.

Sacubitril/Valsartan – Underused?

20 Neprilysin Inhibitors

• Neprilysin: Enzyme which degrades

natriuretic peptides & angiotensin II

• Natriuretic peptide breakdown: Bad
• Angiotensin II breakdown: Good
• OVERTURE Trial
• Omipatrilat – combined neprilysin & ACE

inhibition

• 5770 patients, NYHA Class II-IV
• Randomized: Omipatrilat vs. Enalapril
• Primary endpoint: Death or HF hospitalization

requiring IV diuretics

• Angioedema:
• Omapatrilat: 24 patients (0.8%)
• Enalapril: 14 patients (0.5%)
• Later studies showed >3x risk of angioedema

w/omapatrilat

Adapted from Packer et al. Circulation. 2002;106:920-926.

PARADIGM-HF

• LCZ696 (Neprilysin inhibitor sacubitril with

valsartan) vs. Enalapril 10 bid

• Double-blind, randomized trial of 8442 patients
• LVEF ≤ 40%
• NYHA II-IV – almost all NYHA II-III
• Primary end-point: Time to CV death or HF

hospitalization

• Did not require IV diuretic therapy as in

OVERTURE

• Stopped early after median follow-up of 27

months

• Approved by FDA on July 7, 2015

Adapted from McMurray et al. New Engl J Med. 2014;371:993-1004.

Breakdown of Outcomes

Adapted from McMurray et al. New Engl J Med. 2014;371:993-1004.

Breakdown of Outcomes

Adapted from McMurray et al. New Engl J Med. 2014;371:993-1004.

21

What You Should Be Asking Yourself… Which systolic HF patients should I not be putting on sacubitril/valsartan? Digoxin – Do We Ask Too Much?

Digoxin – DIG Trial (1997)

• 6800 patients with EF ≤ 45%
• Digoxin vs. placebo
• Most patients on diuretics & ACE-

inhibitors

• All patients in sinus rhythm
• Outcomes:
• Primary: All-cause mortality
• Secondary: CV death, worsened HF &

hospitalizations

All-Cause Mortality

Adapted from NEJM. 1997;336:525-33.

22

Death or HF Hospitalization

Adapted from NEJM. 1997;336:525-33.

placebo digoxin

Cochrane Review: Risk of Clinical Deterioration

Adapted from Hood et al. Cochrane Library. 2004, Issue 4.

A Recent Story…

• I was nervously thinking about this talk…

A Recent Story…

23

• I was nervously thinking about this talk…
• Will they like it?

A Recent Story…

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?

A Recent Story…

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.

A Recent Story…

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.
• I took 600 mg of ibuprofen.

A Recent Story…

24

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.
• I took 600 mg of ibuprofen.
• My headache went away. I was happy.

A Recent Story…

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.
• I took 600 mg of ibuprofen.
• My headache went away. I was happy.

A Recent Story…

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.
• I took 600 mg of ibuprofen.
• My headache went away. I was happy.
• Note: I never thought I was going to live

longer by taking the ibuprofen.

A Recent Story…

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.
• I took 600 mg of ibuprofen.
• My headache went away. I was happy.
• Note: I never thought I was going to live

longer by taking the ibuprofen.

• Should I have been denied the opportunity to

take the ibuprofen???

A Recent Story…

25

• I was nervously thinking about this talk…
• Will they like it?
• Will my jokes fall flat?
• I developed a tension headache.
• I took 600 mg of ibuprofen.
• My headache went away. I was happy.
• Note: I never thought I was going to live

longer by taking the ibuprofen.

• Should I have been denied the opportunity to

take the ibuprofen???

• Why do we hold digoxin to such a higher

standard?

A Recent Story…

Statins – Not Always the Answer!

GISSI-HF Trial

• 4574 patients with HF (ischemic or nonischemic)
• NYHA Class II-IV
• EF <40% or EF>40% but HF hospitalization in past 12

months

• Ischemic (40%), Nonischemic (60%)
• Randomized: Rosuvastatin 10 mg daily vs.

placebo

• Primary endpoints:
• Survival
• Mortality or CV hospitalization

Adapted from GISSI-HF Investigators. Lancet. 2008;372:1231-9.

GISSI-HF: Mortality

Adapted from GISSI-HF Investigators. Lancet. 2008;372:1231-9.

26

GISSI-HF: Mortality or CV Hospitalization

Adapted from GISSI-HF Investigators. Lancet. 2008;372:1231-9.

A Mental Exercise A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

27

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large
• Placebo-controlled (not versus alternative lipid-

lowering agent!)

28

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large
• Placebo-controlled (not versus alternative lipid-

lowering agent!)

• Only patients with ischemic cardiomyopathy

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large
• Placebo-controlled (not versus alternative lipid-

lowering agent!)

• Only patients with ischemic cardiomyopathy
• Primary endpoint: Vascular events!

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large
• Placebo-controlled (not versus alternative lipid-

lowering agent!)

• Only patients with ischemic cardiomyopathy
• Primary endpoint: Vascular events!
• Only one problem…

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large
• Placebo-controlled (not versus alternative lipid-

lowering agent!)

• Only patients with ischemic cardiomyopathy
• Primary endpoint: Vascular events!
• Only one problem…
• No way this would ever be considered ethical…

29

A Mental Exercise

• You’re a highly-paid executive at a big

pharmaceutical company which makes an on-patent statin.

• You have been tasked with designing a heart failure

clinical trial to test your statin’s efficacy.

• You are very motivated to get a positive result.
• What type of study would you pitch?
• Large
• Placebo-controlled (not versus alternative lipid-

lowering agent!)

• Only patients with ischemic cardiomyopathy
• Primary endpoint: Vascular events!
• Only one problem…
• No way this would ever be considered ethical…
• Right?

CORONA

• 5011 patients ≥ 60 years
• All with ischemic, systolic HF
• NYHA II: EF ≤ 35%
• NYHA III-IV: EF ≤ 40%
• Rosuvastatin 10 mg daily vs. placebo
• Primary endpoint : A vascular endpoint (!)
• CV death, nonfatal MI, nonfatal stroke

CORONA: Primary Endpoint

Adapted from Kjekshus et al. NEJM. 2007;357:2248-61. Adapted from Kjekshus et al. NEJM. 2007;357:2248-61.

CORONA: All-Cause Mortality

30

So What to Do With Statins in HF?

• For nonischemic heart failure à forget it!
• For ischemic heart failure…
• If angina, PVD, etc. à still use
• Otherwise à consider not using
• Polypharmacy issues are real!
• Let’s focus on what makes a real difference…

Aldosterone Antagonists: Still Underused?

RALES Trial

• 1663 patients with LVEF ≤ 35%
• NYHA Class III-IV
• Had been NYHA Class IV within 6 months of

enrollment

• Excluded if Cr>2.5 mg/dl or K>5.0 mmol/L
• Randomized to spironolactone 25 mg qd vs.

placebo

• Primary outcome: All-cause mortality

RALES Trial: Survival

Adapted from Pitt et al. NEJM. 1999;341:709-17.

31

EMPHASIS-HF Trial

• 2737 patients
• Age >55 years, NYHA II
• EF<30% (or 30-35% with QRS>130 ms)
• Cardiac hospitalization in last 6 months or high

BNP

• Excluded if K > 5.0 or GFR <30 mL/min
• Randomized to eplerenone 50 mg qd vs.

placebo

• Primary endpoint: CV Death or HF

hospitalization

• Stopped early (21 months) due to + result

Adapted from Zannad et al. NEJM. 2011;364:11-21.

EMPHASIS-HF: Primary Endpoint

Adapted from Zannad et al. NEJM. 2011;364:11-21.

EMPHASIS-HF: Overall Mortality

Adapted from Zannad et al. NEJM. 2011;364:11-21.

Ivabradine

• Ivabradine: Inhibitor of If current of S-A node
• No direct effect on myocardial contractility or

intracardiac conduction

• SHIFT
• 6558 patients in SR, resting HR ≥ 70, EF ≤ 35%, and

HF hospitalization within past 12 months.

• Stable standard medical therapy for ≥ 4 weeks
• Randomized: Ivabradine or placebo x mean 23 months
• Primary endpoint: CV Death or HF hospitalization
• Main side-effects:
• Symptomatic bradycardia: 5% vs. 1%
• Visual side-effects (phosphenes): 3% vs. 1%
• Approved by FDA on April 15, 2015

Adapted from Swedberg et al. Lancet. 2010;376:875-885.

32 SHIFT: HR During Study

Adapted from Swedberg et al. Lancet. 2010;376:875-885.

SHIFT: Primary Endpoint

Adapted from Swedberg et al. Lancet. 2010;376:875-885.

Beta-Blocker Doses

Adapted from Swedberg et al. Lancet. 2010;376:875-885.

Percentage of patients (%)

Caution: Beware Unintended Consequences

• Study published in NEJM in 2004
• Examined the impact of the RALES trial
• n spironolactone prescriptions & events in

Canada between 1993-2001 (before/after RALES)

• Primary analysis: Spironolactone

prescriptions & patient outcomes for HF patients on ACE-inhibitors

• Secondary analysis: Spironolactone

prescriptions & patient outcomes for all patients on ACE-inhibitors

33

Spironolactone Prescriptions vs. Time

Adapted from Juurlink et al. NEJM. 2004;351:543-51.

No Change in HF Admissions

Adapted from Juurlink et al. NEJM. 2004;351:543-51.

More Hospitalizations for Hyperkalemia

Adapted from Juurlink et al. NEJM. 2004;351:543-51.

More In-Hospital Hyperkalemic Death

Adapted from Juurlink et al. NEJM. 2004;351:543-51.

34

Summary

• ICDs in heart failure
• Don’t forget them – they can make a

difference

• Ischemic > Nonischemic HF benefit
• Don’t put in immediately post-MI!
• Life-Vest – A role, but a limited one
• Medications
• Carvedilol > Metoprolol
• Use digoxin more
• Use statins less
• Don’t forget about aldosterone antagonists
• If you’re going to use an ACE or ARB...
• Think about sacubitril/valsartan!

A Final Thought: Knowing Trial Data is Good – But Don’t Be Slaves to “The Data” “All generalizations are false – including this one.”

• Mark Twain

A Final Thought: Knowing Trial Data is Good – But Don’t Be Slaves to “The Data”

Thank you!