CRDAC-20201215-UpdatedNovartisSlides.pdf

cardiovascular and renal drug advisory committee meeting.

1. 1
Sacubitril/Valsartan
(ENTRESTO
®
)
Cardiovascular and Renal Drugs Advisory Committee Meeting
Supplemental NDA 207620-S018
Novartis Pharmaceuticals Corporation
December 15, 2020

2. 2
Introduction
David Soergel, MD
Global Head, Cardiovascular, Renal and Metabolism Development
Novartis Pharmaceuticals Corporation

3. 3
To discuss how Entrestocan help address a substantial unmet medical need
in heart failure with preserved ejection fraction (HFpEF)
•
The understanding of HFpEFhas evolved;PARAGON-HF contributes significantly
to this evolution
•Despite the narrowstatistical miss, a true albeit modest treatment effect ofEntresto
is apparentin the PARAGON-HF study population
‒Larger treatment effect observed in patients with lower LVEF and in women
•Evidence fromother trials, including PARADIGM-HF in HFrEF,increase confidence
in the treatment effect of Entresto in HFpEF
•Favorable safety profile in HFpEF, in line with the extensive experience in HFrEF
Why We Are Here Today

4. 4
•Heart failure (HF) is segmented by left ventricular ejection fraction (LVEF)
‒HF with reduced ejection fraction (HFrEF)
‒HF with preserved ejection fraction (HFpEF)
•3.25 million HFpEF patients in the US
•Serious and debilitating disease
‒High patient burden driven by recurrent hospitalizations and reduced quality of life
‒Proportion of HF patients hospitalized with HFpEF is increasing, expected to be 50% in 2020​
‒40%of patients re- admitted within 1 year
•No approved treatment for HFpEF
Heart Failure and HFpEF
<<Virani SS, et al Circulation. 2020;141(9):e139-596; OktayAA, et al. Curr Heart Fail Rep. 2013;10(4):401-410; Cheng RK, et al. Am Heart J 2014;168(5):721 -730>>

5. 5
•Unique dual blocker of renin- angiotensin system (RAS) and neprilysin
•Superiority demonstrated over RAS inhibition alone in HFrEF
‒20% relative risk reduction in HF hospitalization or CV death
•Approved in 115 countries worldwide
•Global exposure>2.6 million patient-years
Entresto
®
(Sacubitril/Valsartan)
Approved for HFrEF in the US since 2015

6. 6
Entresto
®
Registration Program
EOP2: End of Phase 2; sNDA: supplemental New Drug Application
2011201220132014201520162017201820192020
Phase 3
PARADIGM-HF
Completed
HFrEF Approval Pediatric HF
Approval
Phase 2
PARAMOUNT
Completed
Phase 3
PARAGON-HF
Initiated
PARAGON-HF
Completed
FDA Pre-filing
Meetings
sNDA
Submission
HFrEF
HFpEF
FDA EOP2
HFpEF
Meeting

7. 7
Proposed Indication
ENTRESTO is indicated for the treatment of chronic heart failure:
‒To reduce cardiovascular death and hospitalization for heart failure in patients
with chronic heart failure and reduced ejection fraction
‒To reduce worsening heart failure (total heart failure hospitalizations and
urgent heart failure visits) in patients with chronic heart failure and
preserved ejection fractionwith left ventricular ejection fraction
belownormal

8. 8
What You Will Hear Today
•Significant morbidity (recurrent hospitalizations)
•No approved treatment
Unmet Need
•PARAGON-HF
•Additional evidence from PARADIGM-HF and PARAMOUNT
Totality of
Evidence
•Consistent treatment effect across analyses of the primary endpoint and other clinically
relevant endpoints, supporting a true treatment effect
•Greater benefit observed in HFpEF patients with lower LVEF
•Well-characterized and favorable safety profile, consistent with approved indication
Efficacy and
Safety
•Evidence supports extending use to adjacent HFpEF population of patients with LVEF below normal to reduce worsening HF events
Benefit / Risk

9. 9
Introduction David Soergel, MD
Global Head, Cardiovascular, Renal & Metabolism Development
Novartis Pharmaceuticals Corporation
Heart failure with preserved
ejection fraction (HFpEF)John McMurray, MD
Professor, Institute of Cardiovascular & Medical Sciences
BHF Cardiovascular Research Centre
University of Glasgow & Queen Elizabeth University Hospital in Glasgow
Sacubitril/Valsartan
Efficacy and Safety in
HFpEF Scott Solomon, MD
Professor, Harvard Medical School
Brigham and Women’s Hospital
PARAGON- HF:
Clinical Implications John McMurray, MD
Professor, Institute of Cardiovascular & Medical Sciences
BHF Cardiovascular Research Centre
University of Glasgow
& Queen Elizabeth University Hospital in Glasgow
Closing Remarks David Soergel, MD
Global Head, Cardiovascular, Renal & Metabolism Development
Novartis Pharmaceuticals Corporation
Presentation Overview

10. 10
Brian Claggett, PhD Assistant Professor, Harvard Medical School
Brigham and Women’s Hospital
AkshayDesai, MD Associate Professor,Harvard Medical School
Brigham and Woman’s Hospital
G. Michael Felker, MD Professor, Duke University School of Medicine
Duke University Medical Center
Marty Lefkowitz, MD Head, Therapeutic Area Clinical Development
Novartis,Cardiovascular, Renal &Metabolism
Guenther Mueller-Velten, Dipl-Wi-Math Executive Director,Biostatistics
Novartis,Cardiovascular, Renal &Metabolism
Ana de Vera, MD Head, Patient Safety
Novartis,Cardiovascular, Renal & Metabolism
Trish Kay-Mugford, DVM Head, Regulatory Affairs
Novartis,Cardiovascular, Renal &Metabolism
Available for Questions

11. 11
Heart failure with
preserved ejection fraction
(HFpEF)
What is it and why does it matter?
John McMurray, MD
BHF Cardiovascular Research Centre,
University of Glasgow & Queen Elizabeth University Hospital in Glasgow

12. 12
•What is heart failure?
•Left ventricular ejection fraction
•History of heart failure phenotyping by ejection fraction –HFrEF,
HFpEF and HFmrEF
•Clinical characteristics of, and outcomes in, patients with HFrEF,
HFpEF and HFmrEF
•Changing epidemiology of heart failure
•Goals of therapy and treatments available
Outline of Presentation

13. 13
•A clinical syndrome -causing breathlessness (dyspnea), fatigue/exertional limitation and
edema (leg/ankle swelling) caused by dysfunction of the heart (muscle, valves, endocardium,
pericardium, rate or rhythm).
•Common-Afflicts 1-2% of the population (up to 10% of older people)
•Disabling-Greatly reduced quality of life (worse than almost any other chronic condition)
with frequent ED attendances and high rates of hospitalization. Heart failure accounts for 5%
of all medical admissions; is the single most common cause of hospitalization in people >65
years of age.
•Costly-Accounts for 1-2% of health care spending, 70% of which is due to hospitalizations.
•Deadly-High mortality rate ( ~50% at 5 years) but varies according to type of heart failure.
•Progressive-worsens progressively over time, with development of additional
complications including renal dysfunction, anemia, atrial fibrillation and other arrhythmias.
What is Heart Failure?

14. 14
•What is heart failure?
•Left ventricular ejection fraction
•History of heart failure phenotyping by ejection fraction –HFrEF,
HFpEF and HFmrEF
•Clinical characteristics of, and outcomes in, patients with HFrEF,
HFpEF and HFmrEF
•Changing epidemiology of heart failure
•Goals of therapy and treatments available
Outline of Presentation

15. 15
Left Ventricular Ejection Fraction (LVEF)
End-Diastole End-Systole
EDV
ESV LVEF = X100
EDV-ESV
EDV

16. 16
What is a Normal LVEF?

17. 17
Framingham Heart Study: 2005- 2014 (3rd decade)
LVEF distribution in participants without heart failure
Median
Q3
Q1
<<Vasan RS et al JACC Cardiovasc Imaging. 2018;11:1-11>>

18. 18
•What is heart failure?
•Left ventricular ejection fraction
•History of heart failure phenotyping by ejection fraction –
HFrEF, HFpEF and HFmrEF
•Clinical characteristics of, and outcomes in, patients with HFrEF,
HFpEF and HFmrEF
•Changing epidemiology of heart failure
•Goals of therapy and treatments available
Outline of Presentation

19. 19
A Short History of Heart Failure Terminology and
Trials
ACC/AHA
Heartfailurewith
“borderline”
ejectionfraction
(41-49%)
Heartfailure
(no mentionof
ejection
fraction)
CONSENSUS
SOLVD-T
(LVEF ≤35%)
CHARM-
Preserved
(LVEF >40%)
DIG
(LVEF ≤45%;
ancillarytrial>45%)
I-Preserve
(LVEF ≥45%)
V-HeFT
TOPCAT
(LVEF ≥45%)
Heartfailurewith
preserved
ejectionfraction
(>40%)
Heartfailurewith
reducedejection
fraction
ESC
Heartfailurewith
“mid- range”
ejectionfraction
(40-49%)
1987 1991 2006 20162010 2014 201920031997
PARAGON-HF
(LVEF ≥45%)
PARAMOUNT
(LVEF ≥45%)

20. 20
•HFrEF–a disease defined by the results of a trial (based on
an arbitrary LVEF cut-point)
•HFpEF –originally defined as everything else left over i.e.
defined by not having a “reduced” LVEF (<40%)
•HFmrEF–patients with “mid range” or “mildly reduced” LVEF
(
~40-50%) –not normal but not HFrEF; redefined HFpEF as
LVEF >50%. Why 40-50%??
The Heart Failure Syndromes

21. 21
•What is heart failure?
•Left ventricular ejection fraction
•History of heart failure phenotyping by ejection fraction –HFrEF,
HFpEF and HFmrEF
•Clinical characteristics of, and outcomes in, patients with
HFrEF, HFpEF and HFmrEF
•Changing epidemiology of heart failure
•Goals of therapy and treatments available
Outline of Presentation

22. 22
HFrEF, HFmrEF and HFpEF: Similarities &
Differences
Characteristic HFrEF (n=15135) HFmrEF (n=4078) HFpEF (n=9911)
LVEF <40% 40-50% ≥51%
Mean (median) age (yr) 64 (64) 69 (70) 72 (72)
Age >70 years (%) 30 48 57
Female (%) 22 37 57
Medical history (%)
Hypertension 66 79 89
Myocardial infarction 42 41 22
Atrial fibrillation 35 39 39
Diabetes 31 35 36
NYHA class I/II, III/IV (%) 73/27 60/40 55/45
Mean LVEF (%) 29 47 61
Mean systolic BP (mmHg) 122 132 133
Median NT pro BNP (pg/ml) 1420 997 602

23. 23
HFrEF, HFmrEF and HFpEF: Similarities &
Differences
Characteristic HFrEF (n=15135) HFmrEF (n=4078) HFpEF (n=9911)
LVEF <40% 40-50% ≥51%
Mean (median) age (yr) 64 (64) 69 (70) 72 (72)
Age >70 years (%) 30 48 57
Female (%) 22 37 57
Medical history (%)
Hypertension 66 79 89
Myocardial infarction 42 41 22
Atrial fibrillation 35 39 39
Diabetes 31 35 36
NYHA class I/II, III/IV (%) 73/27 60/40 55/45
Mean LVEF (%) 29 47 61
Mean systolic BP (mmHg) 122 132 133
Median NT pro BNP (pg/ml) 1420 997 602

24. 24
7,599 patients with NYHA class II-IV heart failure and no LVEF exclusion
Female
Male
LVEF in Men and Women with Heart Failure
LVEF distribution in CHARM
<<McMurray JJV, Jackson AM et al Circulation. 2020;141:338–351>>

25. 25
HFrEF, HFmrEF and HFpEF: Same Symptoms
and Signs
Essentially all patients limited by dyspnea on exertion, irrespective of LVEF category
PND –Paroxysmal nocturnal dyspnea; JVD –Jugular venous distension; S3 –Third heart sound
p-values (trend) for all differences <0.001 except rales
Dyspnea missing in IPreserve & TOPCAT.
0
10
20
30
40
50
%
HFrEF HFmrEF HFpEF
Dyspnea on effort
HFrEF 85.4%
HFmrEF 94.4%
HFpEF 93.7%

26. 26
0
10
20
30
40
50
60
70
80
90
100
Clinical Summary Score Overall Summary Score Total Symptom Score
KCCQ Summary Score *
HFrEF HFmrEF HFpEF
HFrEF, HFmrEF and HFpEF: Health- related
Quality of Life
*Unadjusted mean scores
Trials analysed –ATMOSPHERE, PARADIGM -HF, PARAGON-HF & TOPCAT-Americas
p-values for all differences <0.001
Maximum possible score 100
Lower score = worse HRQL

27. 27
HFrEF, HFmrEF and HFpEF: Hospitalization
Figures truncated to 5 years but event rates based on entire follow-up period
Total HF hospitalizations investigator reported except for PARAGON-HF.
HFrEF –ATMOSPHERE & PARADIGM -HF
HFpEF –CHARM-Preserved, IPreserve, PARAGON-HF & TOPCAT-Americas
Similar rates of heart failure hospitalization

28. 28
HFrEF, HFmrEF and HFpEF: Mortality
HFrEF –ATMOSPHERE & PARADIGM -HF
HFpEF –CHARM-Preserved, IPreserve, PARAGON-HF & TOPCAT-Americas
Figures truncated to 5 years but event rates based on entire follow-up period

29. HFrEF HFmrEF
HFpEF
HFrEF, HFmrEF and HFpEF: Causes of Death
Sudden death
Worsening HF
Other CV causes
Non-CV/unknown
16%
35%
22%
27%
17%
20%
25%
38%
28% 30%
16%26%
Any CV death = 84% Any CV death = 72%
Any CV death = 62%
29

30. 30
•What is heart failure?
•Left ventricular ejection fraction
•History of heart failure phenotyping by ejection fraction –HFrEF,
HFpEF and HFmrEF
•Clinical characteristics of, and outcomes in, patients with HFrEF,
HFpEF and HFmrEF
•Changing epidemiology of heart failure
•Goals of therapy and treatments available
Outline of Presentation

31. 31
Increasing Prevalence of HF in the USA
<<AHA Policy Statement: Forecasting the Impact of Heart Failure in the United States, Circ Heart Fail. 2013;6:606-619>>

32. 32
Patients hospitalized with heart failure (“Get With The Guidelines”)
P
trend<0.0001
EF ≥50%
EF <40%
EF ≥40 <50%
Changing HF Phenotype Over Time
<<Steinberg BA et al Circulation 2012; 126: 65-75>>
33 35 36 35 38 39
15 14 14 14
13
14
52 51 50 51 49 47

33. 33
Projected Cost of HF in the USA
<<AHA Policy Statement: Forecasting the Impact of Heart Failure in the United States, Circ Heart Fail. 2013;6:606-619>>

34. 34
•What is heart failure?
•Left ventricular ejection fraction
•History of heart failure phenotyping by ejection fraction –HFrEF,
HFpEF and HFmrEF
•Clinical characteristics of, and outcomes in, patients with HFrEF,
HFpEF and HFmrEF
•Changing epidemiology of heart failure
•Goals of therapy and treatments available
Outline of Presentation

35. 35
Overarching goal – slow progressive worsening over time
•Reduce rate of deterioration in symptoms and quality of life (or
even improve symptoms) –
physician evaluation (NYHA class), patient
self-assessment (KCCQ)
•Reduce episodes of worsening e.g. leading to ED visits and
hospital admission (and re- admissions) –
time-to-first event,
recurrent/repeat events
•Reduce mortality, where possible – generally only cardiovascular
mortality modifiable
Goals of Therapy

36. 36
•ACE inhibitors and ARBs
•Digoxin
•Beta-blockers
•MRAs
•Ivabradine
•Sacubitril/valsartan
•SGLT2 inhibitors
•Vericiguat?
•Devices: ICD, CRT
Treatments for Heart Failure
HFrEF HFpEF

37. 37
Effects of Candesartan and Spironolactone
According to LVEF
Because of the growing
interest in HFmrEF
(“mid-range” or
“mildly-reduced” EF)

38. 38
CV death or HF hospitalizations (time-to-first)
Continuous HR
placebo better
95% CI
active better
Continuous HR
Candesartan and Spironolactone: Treatment
Effect by LVEF
Adapted from Dewan P et al Eur J Heart Fail 2020;22: 898-901

39. 39
•Multiple effective therapies are available for patients with HFrEF, presently defined as a LVEF <40%
•The remaining patients with heart failure were originally described as having HFpEF, although,
recently, HFpEF has been re- defined –people with a LVEF 40- 50% have been categorized as
having HFmrEF and only those with a LVEF >50% now described as having HFpEF (arbitrary cut-
points)
•No treatments are approved for people with HFpEF and HFmrEF, although they account for 50% of
cases of heart failure and are increasing in incidence and prevalence (as the predominant
phenotypes in aging populations)
•Although patients HFpEF and HFmrEF have a lower mortality rate (and proportionately more non-
cardiovascular deaths) than patients with HFrEF, their rates of HF hospitalization are nearly as high
as in patients with HFrEF
•People with HFpEF and HFmrEF have at least as many symptoms and as great a reduction in
quality of life as patients with HFrEF
•Heart failure with LVEF ≥40% is an important unmet treatment need
Summary

40. 40
Sacubitril/Valsartan Efficacy and
Safety in HFpEF
Scott Solomon, MD
Professor, Harvard Medical School
Brigham and Women’s Hospital

41. 41<<VardenyO. JACC-HF 2014>>
Sacubitril/Valsartan: A First-in-Class Angiotensin
Receptor Neprilysin Inhibitor
Sacubitril/Valsartan
ValsartanSacubitril
Sacubitrilat
NH
N
N
N
N
O
OH
O
Heart Failure Renin
AngiotensinSystem
Vasoactive
PeptideSystem
pro-BNP
NT-pro BNP
ANP BNP CNP
Adrenomedullin
Bradykinin
Substance P
(angiotensin II)
AT
1receptor
Angiotensinogen
(liver secretion)
Angiotensin I
Angiotensin II
Vasoconstriction
blood pressure
sympathetic tone
aldosterone
fibrosis
hypertrophy
Vasodilation
blood pressure
sympathetic tone
aldosterone levels
fibrosis
hypertrophy
Natriuresis/Diuresis
Inactive
fragments
OH
O
NH
O
OH
O
Novel crystalline complex
consisting of the molecular
moieties of valsartan and
sacubitril in an equimolar ratio
Neprilysin

42. 42
Sacubitril/Valsartan Efficacy Results in HFrEF
PARADIGM-HF -Significant Reduction in Primary Endpoint, CV Death and All-Cause Mortality
Primary Endpoint (CV death
and HF hospitalization)
0
16
32
40
24
8
Enalapril
(n=4212)
Sacubitril/
Valsartan
(n=4187)
HR = 0.80 (0.73-0.87)
P = 0.0000004
Number needed to treat = 21
360 720 10800180 540 900 1260
Kaplan-Meier Estimate of
Cumulative Rates (%)
Days After Randomization
1117
914
4187
4212
3922
3883
3663
3579
3018
2922
2257
2123
1544
1488
896
853
249
236
Patients at Risk
Sac/Val
Enalapril
835
711
All-cause Mortality
Enalapril
(n=4212)
Sacubitril/
Valsartan
(n=4187)
HR = 0.84 (0.76-0.93)
P<0.0001
360 720 10800180 540 900 1260
0
16
32
24
8
Kaplan-Meier Estimate of
Cumulative Rates (%)
Days After Randomization
4187
4212
4056
4051
3891
3860
3282
3231
2478
2410
1716
1726
1005
994
280
279
Patients at Risk
Sac/Val
Enalapril
CV Death
Enalapril
(n=4212)
Sacubitril/
Valsartan
(n=4187)
HR = 0.80 (0.71-0.89)
P = 0.00008
Number need to treat = 32
360 720 10800180 540 900 1260
0
16
32
24
8
693
558
Kaplan-Meier Estimate of
Cumulative Rates (%)
Days After Randomization
4187
4212
4056
4051
3891
3860
3282
3231
2478
2410
1716
1726
1005
994
280
279
Patients at Risk
Sac/Val
Enalapril
<<McMurray et al. NEJM 2014 371:993-1004>>

43. 43
Phase II Results in HFpEF
PARAMOUNT -Improvement in Several Domains
NT-proBNP: N-terminal pro- brainnatriureticpeptide; NYHA: New York Heart Association
12-week randomized, double-blind, active-controlled study evaluating sacubitril/valsartan compared to valsartan
on changes in NT-proBNP, cardiac structure and function, and HF symptoms/signs in 301 HFpEF patients
<<Solomon et al. Lancet2012 380(9851):1387-1395>>
NT-proBNP(pg/mL)
Week 36
P=0.20
Week 12
P=0.005
Week 4
P=0.063
1000
200
300
400
500
600
700
800
900
Weeks post-randomization
0 5 10152025303540
Entresto
Valsartan
Improvement in NT-proBNP Improvement in NYHA class
Percent of patients
100
0
20
30
40
50
60
70
80
90
10
P=0.11
P=0.05
Week 12 Week 36
Sac/ValValsartan Sac/ValValsartan
Worsened
Unchanged Improved
Improvement in left atrial size
P=0.18
P=0.003
36 Weeks12 Weeks
2
-6
-5
-4
-3
-2
-1
0
1
Entresto
Valsartan
Change in left atrial volume (mL)

44. 44
<<Solomon SD, et al. JACC-Heart Fail 2017; 5(7):471-482. >>
Randomization 1:1
~35 months
Valsartan 160 mg BID*
Sacubitril/Valsartan 97/103 mg BID*
On top of optimal background medications for co-
morbidities(excluding ACEi and ARB)
Double-blind treatment period
1 –4 weeks
Sacubitril/
Valsartan
49/51 mg BID
Valsartan
80 mg BID
Eligibility
Screening
2 –4 weeks
Active single-blind run -in
period
Valsartan
40 mg BID
PARAGON- HF: Trial Design
Randomized, double- blind, active comparator trial testing the hypothesis that sacubitril/valsartan, compared with
valsartan, would reduce the composite outcome of total HF hospitalizations and CV death
* target doses, down- titration allowed for tolerability

45. 45
PARAGON- HF: Endpoints
Composite Primary Endpoint
•Composite of CEC-confirmed:
‒Total (first and recurrent) HF
hospitalization
‒CV death
Secondary Endpoints
•NYHA functional classification at 8 months
•KCCQ Clinical Summary Score at 8 months
•Time to first occurrence of worsening renal function
•Time to all-cause mortality

46. 46
PARAGON- HF: Endpoints
Expanded Composite Endpoint
(Exploratory)
•Composite of CEC-confirmed:
‒Total (first and recurrent) HF hospitalization
‒CV death
‒Urgent HF visits
Composite Primary Endpoint
•Composite of CEC-confirmed:
‒Total (first and recurrent) HF
hospitalization
‒CV death
Secondary Endpoints
•NYHA functional classification at 8 months
•KCCQ Clinical Summary Score at 8 months
•Time to first occurrence of worsening renal function
•Time to all-cause mortality

47. 47
Valsartan chosen as the active comparator to treat comorbidities
•~85% of patients treated with ACEi/ARB in prior HFpEF trials and prior to
PARAGON run-in
•Similar RAS blockade in both groups; sacubitril/valsartan 97/103 mg provides
similar plasma exposure as valsartan 160 mg
•Allowed assessment of incremental effect of sacubitril on RAS inhibition
•Of note, previous studies suggests a modest benefit from RAS inhibition
Key Design Considerations: Active Comparator
<<Yusuf et al. Lancet 2003;362:777– 81; Rogers et al. European Journal of Heart Failure 2014;16:33– 40.>>

48. 48
Key Design Considerations: Recurrent Events
•HFpEF is characterized by frequent recurrent worsening HF events (HF hospitalization and Urgent HF visits)
•Each event is associated with worsening of long term prognosis
‒In the CHARM program (candesartan) the risk of death increased with each additional HF hospitalization, with an
~30% cumulative incremental risk associated with 2nd or 3rd HF hospitalization
•Traditional TTFE analysis ignore all events after the first event
•This approach more accurately reflects the true burden of the illness on the patient and the healthcare system
•Commonly used in other diseases where recurrent encounters are common (e.g. asthma, multiple sclerosis) and included in
June 2019 FDA guidance “Treatment for Heart Failure: Endpoints for Drug Development”
Acutely
decompensated
Compensated
Chronically
decompensated
Clinical status
Death
Recurrent
Worsening
Heart Failure
Event
Disease progression
<<Rogers et al. 2013>>

49. 49
Key Inclusion & Exclusion Criteria
To Avoid Overlap with HFrEF Population and Ensure Certainty of HFpEF Diagnosis
Key inclusion criteria Key exclusion criteria
•≥ 50 years of age and LVEF ≥ 45%
•Heart failure signs/symptoms (NYHA Class II–IV)
requiring treatment with diuretic(s) for at least 30
days prior to enrollment
•Structural heart disease (LAE or LVH by
echocardiography)
•Elevation in natriuretic peptides
•NT-proBNP 200 pg/mlif hospitalized for HF
within 9 months, and 300 pg/ml if not
hospitalized; 3-fold increase for patients in
AF at enrollment
•Any prior measurement of LVEF < 40%
•Current acute decompensated heart failure
•Alternative reason for signs and symptoms
•SBP < 110 or ≥ 180mm Hg (or > 150mm Hg if
patient not taking 3 or more antihypertensive
medications)
<<Solomon SD, et al. JACC-Heart Fail 2017>>
LAE: Left Atrial Enlargement; LVH: Left Ventricular Hypertrophy; SBP: Systolic Blood Pressure

50. 50
Clinical Endpoint Committee (CEC) Adjudication
Criteria for HF hospitalization and Urgent HF Visit
Strict CEC definition for adjudicating HF hospitalizations and Urgent HF visits, consistent with
the Standardized Data Collection for Cardiovascular Trials Initiative (SCTI) (Hicks et al 2018),
all of the below required from source documentation:
Criteria HF Hospitalization
Unplanned presentation with HF exacerbation 
Stay traversing change in calendar date 
At least one symptom of worsening HF 
At least two signs of worsening HF 
Qualified treatments directed at treating HF 
<<Hicks et al. Circulation 2018 137(9):961-972>>

51. 51
Clinical Endpoint Committee (CEC) Adjudication
Criteria for HF hospitalization and Urgent HF Visit
Strict CEC definition for adjudicating HF hospitalizations and Urgent HF visits, consistent with
the Standardized Data Collection for Cardiovascular Trials Initiative (SCTI) (Hicks et al 2018),
all of the below required from source documentation:
<<Hicks et al. Circulation 2018 137(9):961-972>>
Criteria HF HospitalizationUrgent HF Visit
Unplanned presentation with HF exacerbation  
Stay traversing change in calendar date  Not required
At least one symptom of worsening HF  
At least two signs of worsening HF  
Qualified treatments directed at treating HF  + must include IV HF
treatment

52. 52
PARAGON- HF was a Global Trial
848 Sites in 43 Countries
559
370
1327
762
48079
11010960473212
388254204123102654241383614128
1804
3633102522411801701106360231913
26583797970
6148332420

53. 53
Patient Disposition: High Completion Rate
Screened 10,359 in 848 sites in 43 countries
5,746 entered the valsartan run- in
5,205 entered the sacubitril/valsartan run- in
Final vital status known, n = 2,385
Final vital status unknown, n = 4*
Final vital status known n = 2,402
Final vital status unknown n = 5*
Excluded from FAS because of site
closure due to GCP violation
Sacubitril/
Valsartan
N = 12
Valsartan
N = 14
Sacubitril/Valsartan run-in failures
n = 384 (7.4%)
N = 4,796
Valsartan run-in failures
n = 541 (9.4%)
4,822 Randomized
Median follow-up 35 months
Only 7 Patients Withdrew Consent and 2 Patients Lost to Follow-up at End of Study
Sacubitril/Valsartan 200 mg BID
N = 2,407
Valsartan 160 mg BID
N = 2,389
*
7 withdrew consent, 2 lost to follow-up
<<Solomon SD et al. N Engl J Med 2019;381:1609-1620 >>

54. 54
Baseline Demographics
Sacubitril/Valsartan
N=2,407
Valsartan
N=2,389
Age (years) –mean (SD) 72.7 (8.3) 72.8 (8.5)
Sex Male 48% 48%
Female 52% 52%
Race Caucasian 82% 81%
Black 2% 2%
Asian 12% 13%
Region North America* 12% 11%
Latin America 8% 8%
Western Europe 29% 29%
Central Europe 36% 36%
Asia/Pacific/other** 16% 16%
Baseline LVEF –median [IQR] 57 [51,62] 57 [50,63]
Baseline NT-proBNP (pg/mL) –median (IQR) –Sinus rhythm 611 [389, 1072] 583 [370, 1046]
Baseline NT-proBNP (pg/mL) –median (IQR) –Atrial fibrillation 1536 [1153, 2212] 1633 [1191, 2368]
*North America = US and Canada. **Asia/Pacific/Other includes Israel, South Africa, Australia, China, India, Japan, Rep of Korea, Philippines, Singapore, Taiwan.
<<Solomon SD et al. N Engl J Med 2019;381:1609-1620 >>
Well Balanced and Geographically Diverse

55. 55
Sacubitril/Valsartan
N=2,407
Valsartan
N=2,389
NYHA class at randomizationClassI 3% 3%
Class II 78% 77%
Class III 19% 20%
Class IV 0.3% 0.5%
BMI –mean (SD) 30.2 (4.9) 30.3 (5.1)
Baselinemean SBP (SD) / mean DBP (SD) at randomization 130.5 (15.6) / 74.3 (10.6)130.6 (15.3) / 74.3 (10.4)
Medical history Hypertension 96% 95%
Diabetes mellitus 43% 43%
Atrial fibrillation or flutter at screening ECG32% 33%
Hospitalization for HF within 9 months 38% 39%
Medications
At screening ACEi or ARBs 86% 86%
At randomizationDiuretics 95% 96%
MRA 25% 27%*
Beta blockers 80% 79%
Calcium channel blockers 34% 34%
Baselinecharacteristics balanced if not noted by *P<0.05.
<<Solomon SD et al. N Engl J Med 2019; 381:1609-1620 >>
Balanced Baseline Cardiovascular Profile
Baseline Demographics

56. 56
PARAGON- HF Primary Composite Endpoint Results
*Semiparametric LWYY method
Recurrent Event Analysis of Total HF Hospitalizations and CV Death*
Valsartan (n = 2389)
1009 events, 14.6 per 100 pt-years
Sacubitril/Valsartan (n = 2407) 894 events, 12.8 per 100 pt-years
Rate ratio 0.87 (95% CI 0.75, 1.01)
P= 0.059
Total HF hospitalizations and CV death
Mean cumulative events per 100 patients
Years
55
0
5
10
15
20
25
30
35
40
45
50
0 1 2 3 4

57. 57
HF Hospitalizations and CV Death
*HF Hospitalization = Joint frailty model
**CV death = Cox’s proportional hazard model
Mean cumulative events
per 100 patients
55
0
0
10
15
20
25
30
35
40
45
50
5
1 2 3 4
HF hospitalizations*
Events
Valsartan 797
Sacubitril/Valsartan690
Rate ratio 0.85 (95% CI 0.72, 1.00)
Years
Proportion
0.55
0.00
0
0.10 0.15 0.20 0.25 0.30
0.35
0.40
0.45
0.50
0.05
1 2 3 4
CV Death**
Years
Patients
Valsartan 212 (8.9%)
Sacubitril/Valsartan204 (8.5%)
Hazard ratio 0.95 (95% CI 0.79, 1.16)

58. 58
•Expanded Composite Endpoint
‒Prespecified exploratory endpoint adding CEC-confirmed urgent HF visits
•Investigator reported events
‒Prespecified analysis using total events reported by investigators in the trial
•Re-adjudication of CEC-unconfirmed HF hospitalization events
‒Implemented following recommendation from FDA
‒Intended to address concerns that strict adjudication criteria may have discarded a number of
true HF hospitalization events
Analyses Supporting Efficacy on Reducing
Morbidity of HFpEF

59. 59
•Urgent, non- hospitalized, heart failure visits have similar prognostic and
discriminative ability as heart failure hospitalizations
•Incorporated into composite endpoint of recent heart failure trials (MADIT-CRT,
DAPA- HF) and included in June 2019 FDA guidance “Treatment for Heart Failure:
Endpoints for Drug Development”
‒Of 1582 worsening HF events, 95 (6.0%) were urgent HF visits.
•Patients whose first episode of worsening HF event was an urgent visit had
similar age, comorbidities, baseline NT-proBNP, and MAGGIC risk scores to
those in whom the first HF event was a hospitalization.
Expanded Composite Endpoint
CV Death, Total HF Hospitalizations and Urgent HF Visits
CEC-Adjudicated Pre- Specified Exploratory Endpoint

60. 60
Expanded Composite Endpoint
CV Death, Total HF Hospitalizations and Urgent HF Visits
Prospectively CEC-adjudicated Pre- Specified Exploratory Endpoint
Sacubitril/Valsartan
N=2407
Valsartan
N=2389
Rate Ratio/
Hazard Ratio
(95%CI)
P-Value
(2-sided)
nRate per 100 pt yr
(95% CI)
n Rate per 100 pt yr
(95% CI)
CEC-confirmed
Primary Endpoint
894
12.8
(12.0 –13.7)
1009
14.6
(13.7 –15.6)
0.87
(0.75 –1.01)
0.059
Expanded Composite
Endpoint
934
13.4
(12.6 –14.3)
1064
15.4
(14.5 –16.4)
0.86
(0.75 –0.99)
0.040
Total worsening HF*
730
10.5
(9.7 –11.3)
852
12.4
(11.5 –13.2)
0.84
(0.71 –0.98)
0.031
* Total worsening HF events include HF hospitalizations and urgent HF visits
Includes 95 additional events (40 events in sacubitril/valsartan and 55 events in valsartan)

61. 61
Primary Composite Endpoint: Investigator-Reported
Sacubitril/Valsartan
N=2407
Valsartan
N=2389
Rate Ratio/
Hazard Ratio
(95%CI)
P-Value
(2-sided)
nRate per 100 pt yr
(95% CI)
n Rate per 100 pt yr
(95% CI)
CEC-confirmed
Primary Endpoint
894
12.8
(12.0 –13.7)
1009
14.6
(13.7 –15.6)
0.87
(0.75 –1.01)
0.059
Investigator-Reported
Primary Endpoint
1064 15.3
(14.4 –16.2)
1241 18.0
(17.0- 19.0)
0.84
(0.74 –0.97)
0.014
Total HeartFailure
Hospitalizations
916 13.2
(12.3 –14.0)
1087 15.8
(14.8 –16.7)
0.82
(0.71 –0.96)
0.010
Cardiovascular
Death
148 2.1
(1.8 –2.5)
154 2.2
(1.9 –2.6)
0.95
(0.76 –1.19)
0.665
Includes a net of 402 additional events (170 events in sacubitril/valsartan and 232 events valsartan)

62. 62
Clinical Endpoint Committee (CEC) Adjudication
Criteria for HF hospitalization
Strict CEC definition for adjudicating HF hospitalizations consistent with the Standardized Data
Collection for Cardiovascular Trials Initiative (SCTI) (Hicks et al 2018), all of the below required
from source documentation:
HF hospitalization
Unplanned presentation with HF exacerbation
Stay traversing change in calendar date
At least one symptom of worseningHF
At least two signs of worseningHF
Qualified treatments directed at treating HF
<<Hicks et al. Circulation 2018 137(9):961-972>>

63. 63
While stringent adjudication criteria increased specificity, it may have discarded a
number of true HF hospitalization events that did not meet the strict definition,
thereby reducing sensitivity
FDA recommended an independent panel to re-adjudicate unconfirmed HF
hospitalizations:
•The trigger was not concern regarding the quality of the initial CEC adjudication but that
the strict definition used reduced the sensitivity
•Consisted of 3 independent, blinded HF experts
•Reviewed 566 investigator-reported but CEC unconfirmed HF hospitalizations
•Each expert ascribed a probability of being a true HF hospitalization based on clinical
judgement using CEC endpoint packages (100%, 75%, 50%, 25%, 0%)
•The 3 probabilities were averaged for each event and used in a multiple imputation
approach to include re-adjudicated events in the primary analysis
FDA Recommended Re- Adjudication Panel

64. 64
Primary Endpoint Incorporating CEC-confirmed and
Re-adjudicated Events
Average no. of events/N
EAR
Sacubitril/Valsartan vs.
Valsartan
Endpoint
Sacubitril/
Valsartan
N=2407
Valsartan
N=2389
Rate ratio
(95% CI)
2-sided
P-value
CEC-confirmed Primary
Endpoint 894
(12.8)
1009
(14.6)
0.87
(0.75 –1.01)
0.059
Re-adjudication Analysis
Primary Endpoint
999
(14.3)
1135
(16.5)
0.86
(0.75 –1.00)
0.043
Total HF hospitalizations
795
(11.4)
923
(13.4)
0.84
(0.72–0.99)
0.037
EAR: exposure- adjustedrate
Includes equivalent of 231 additional HF Hospitalizations (105 events in sacubitril/valsartan and 126 events in valsartan)

65. 65
Summary of Primary and Expanded Composite
Endpoint Results
Endpoint Sac/Val
N=2407
Valsartan
N=2389
Rate Ratio
(95% CI)
Rate Ratio (95% CI) 2-sided
P-value
CEC-confirmed
Primary endpoint 894 1009 0.870 (0.753, 1.005) 0.059
Expanded composite endpoint*934 10640.861 (0.747, 0.993) 0.040
Supportive analyses of the primary endpoint
Investigator reported
primary endpoint events
1064 12410.843 (0.736, 0.966) 0.014
Investigator reported expanded
composite endpoint**
1200 14140.834 (0.733, 0.950) 0.006
Re-adjudication analysis
of primary endpoint events
999 11350.865 (0.751, 0.995) 0.043
* Exploratory endpoint including CV death and total HF hospitalizations + urgent HF visits
** Investigator reported CV death and total HF hospitalizations + urgent HF visits
0.4 10.6 0.8 1.21.4
sacubitril/valsartan bettervalsartan better

66. 66
SecondaryOutcomes

67. 67
Change in NYHA Functional Class and KCCQ CSS
Favors Sacubitril/Valsartan
Kansas City Cardiomyopathy Questionnaire (KCCQ) Clinical Summary Score (CSS) based on the physical limitation and total symptom score domains
LSM: least square mean; OR: odds ratio
*Patients who died were included in the worsened category
Sac/Val Valsartan
Effect Size
(95% CI)
P-value
(2-sided)
NYHA class favorable change at 8 months – n 2316 2302
Improved
Unchanged
Worsened*
15.0%
76.3%
8.7%
12.6%
77.9%
9.6%
OR, 1.45(1.13 –1.86) 0.004
KCCQ clinical summary score at 8 months – n 2250 2226
LSM of changefrom baseline(SE) -1.5 (0.4) -2.5 (0.4)Difference, 1.0 (0.0 –2.1) 0.051
≥5 point Improvement 33.0% 29.6% 0.019
≥5 point Deterioration 33.5% 34.5% 0.467

68. 68
Composite Renal Endpoint
Sacubitril/Valsartan Reduced the Risk of the Composite Renal Endpoint
Primarily Driven by Lower Incidence of ≥50% Reduction in eGFR
Similar effect seen in PARADIGM-HF with lower risk of the same renal composite endpoint with sacubitril/valsartan compared with enalapril
Composite
Sacubitril/
Valsartan
N=2407
n(%)
Valsartan
N=2389
n(%)
Composite renal endpoint 33
(1.4)
64
(2.7)
(i) Renaldeath 1 1
(ii) Reaching ESRD 7 12
(iii)≥50% decline in
eGFR relative to
baseline
27 60
0.10
0.00
0.02
0.04
0.06
0.08
0 1 2 3 4
2389
2407
2273
2320
2145
2195
1033
1049
135
129
Number at Risk
valsartan
sacubitril/valsartan
Proportion of Patients
Years
HR 0.50 (95% Cl 0.33 –0.77), p = 0.001
valsartan
sacubritril/valsartan

69. 69
0.4 0.8 1.6
0.4 0.8 1.6
Pre-Specified Subgroups for Primary Endpoint
Subgroup
No. of events
/patients
Rate ratio
(95% CI)
Overall 1903/4796 0.87 (0.75−1.01)
Age (years)
Less than 65 years276/825 0.99 (0.64−1.53)
65 years or older1627/3971 0.85 (0.73−0.99)
Age (years)
Less than 75 years938/2597 0.82 (0.66−1.02)
75 years or older 965/2199 0.92 (0.76−1.11)
Sex*
Male 980/2317 1.03 (0.85−1.25)
Female 923/2479 0.73 (0.59−0.90)
Race
Caucasian 1542/3907 0.83 (0.71−0.97)
Black 89/102 0.69 (0.24−1.99)
Asian 237/607 1.25 (0.87−1.79)
Other 35/180 1.03 (0.47−2.28)
Region
North America 478/559 0.80 (0.57−1.14)
Latin America 83/370 1.33 (0.75−2.36)
Western Europe 544/1390 0.69 (0.53−0.89)
Central Europe 466/1715 0.97 (0.76−1.24)
Asia/Pacific 332/762 1.10 (0.79−1.52)
Diabetic
Yes 1041/2069 0.89 (0.74−1.09)
No 862/2727 0.84 (0.68−1.03)
LVEF*
at or below median (57%) 1048/2495 0.78 (0.64−0.95)
above median (57%) 855/2301 1.00 (0.81−1.23)
History of AF
Yes 1140/2521 0.83 (0.69−1.00)
No 763/2275 0.94 (0.75−1.18)
Screening NT−proBNP
at or below median (911
pg/mL)
708/2379
0.85 (0.67−1.08)
above median (911 pg/mL) 1183/2378 0.87 (0.73−1.05)
Screening SBP
at or below median (137 mmHg)
984/2450
0.88 (0.72−1.07)
above median (137 mmHg) 919/2344 0.86 (0.69−1.06)
MRA use
Yes 545/1239 0.73 (0.56−0.95)
No 1358/3557 0.94 (0.79−1.12)
Baseline eGFR
<60 mL/min/1.73m
2 1115/2341 0.79 (0.66−0.95)
>=60 mL/min/1.73m
2 787/2454 1.01 (0.80−1.27)NYHA class
I/II 1402/3843 0.90 (0.76−1.06)
III/IV 499/951 0.79 (0.59−1.06)
Subgroup
No. of events
/patients
Rate ratio
(95% CI)
Rate ratio (95% CI) Rate ratio (95% CI)
*Multivariate interaction P< 0.05.
Evidence for Heterogeneity for Two Subgroups in Multivariable Analysis
<<Solomon SD et al. N Engl J Med; 2019 >>

70. 70
0.4 0.8 1.6
0.4 0.8 1.6
Pre-Specified Subgroups for Primary Endpoint
Subgroup
No. of events
/patients
Rate ratio
(95% CI)
Overall 1903/4796 0.87 (0.75−1.01)
Age (years)
Less than 65 years276/825 0.99 (0.64−1.53)
65 years or older1627/3971 0.85 (0.73−0.99)
Age (years)
Less than 75 years938/2597 0.82 (0.66−1.02)
75 years or older 965/2199 0.92 (0.76−1.11)
Sex*
Male 980/2317 1.03 (0.85−1.25)
Female 923/2479 0.73 (0.59−0.90)
Race
Caucasian 1542/3907 0.83 (0.71−0.97)
Black 89/102 0.69 (0.24−1.99)
Asian 237/607 1.25 (0.87−1.79)
Other 35/180 1.03 (0.47−2.28)
Region
North America 478/559 0.80 (0.57−1.14)
Latin America 83/370 1.33 (0.75−2.36)
Western Europe 544/1390 0.69 (0.53−0.89)
Central Europe 466/1715 0.97 (0.76−1.24)
Asia/Pacific 332/762 1.10 (0.79−1.52)
Diabetic
Yes 1041/2069 0.89 (0.74−1.09)
No 862/2727 0.84 (0.68−1.03)
LVEF*
at or below median (57%) 1048/2495 0.78 (0.64−0.95)
above median (57%) 855/2301 1.00 (0.81−1.23)
History of AF
Yes 1140/2521 0.83 (0.69−1.00)
No 763/2275 0.94 (0.75−1.18)
Screening NT−proBNP
at or below median (911
pg/mL)
708/2379
0.85 (0.67−1.08)
above median (911 pg/mL) 1183/2378 0.87 (0.73−1.05)
Screening SBP
at or below median (137 mmHg)
984/2450
0.88 (0.72−1.07)
above median (137 mmHg) 919/2344 0.86 (0.69−1.06)
MRA use
Yes 545/1239 0.73 (0.56−0.95)
No 1358/3557 0.94 (0.79−1.12)
Baseline eGFR
<60 mL/min/1.73m
2 1115/2341 0.79 (0.66−0.95)
>=60 mL/min/1.73m
2 787/2454 1.01 (0.80−1.27)NYHA class
I/II 1402/3843 0.90 (0.76−1.06)
III/IV 499/951 0.79 (0.59−1.06)
Subgroup
No. of events
/patients
Rate ratio
(95% CI)
Rate ratio (95% CI) Rate ratio (95% CI)
*Multivariate interaction P< 0.05.
Evidence for Heterogeneity for Two Subgroups in Multivariable Analysis
<<Solomon SD et al. N Engl J Med; 2019 >>

71. 71
Significant Heterogeneity in Multivariate Analysis Driven
by Ejection Fraction and Sex
Primary endpoint
Male 980/23171.03 (0.85– 1.25)
0.73 (0.59– 0.90)
Sex
Female 923/2479
at or below median (57%) 1048/24950.78 (0.64– 0.95)
1.00 (0.81– 1.23)
LVEF
above median (57%) 855/2301
Rate ratio (95% CI)
P= 0.035 (categorical)
P= 0.004 (continuous)
P= 0.002 (categorical)
P= 0.005 (continuous)
Multivariable
interactionP-value
Rate ratio
(95% CI)
No. of events/
patients
Subgroup
Only Interactions for Sex and Ejection Fraction Remained Nominally Significant
<<Solomon SD et al. N Engl J Med; 2019 >>
0.4 10.6 0.8 1.21.4

72. 72
Subgroup
No. of
Events/Patients
Rate Ratio
(95% CI)
Overall EF 1903/4796 0.87 (0.75– 1.01)
≤50 512/1208 0.82 (0.63– 1.06)
>50–57 536/1287 0.77 (0.57– 1.03)
>57–63 467/1202 0.91 (0.68– 1.22)
>63 388/1099 1.09 (0.80– 1.47)
Treatment Effect by Ejection Fraction Quartiles
Primary Composite Total HF Hospitalizations and CV Death
Subgroup estimates are based on re- running the primary analysis model for each of the four subgroups defined by the by LVEF quartiles.
0.4 1.610.6 0.8 1.21.4

73. 73* Total worsening HF events include HF hospitalizations and urgent HF visits
Treatment Effect for HFpEF Patients with
LVEF ≤ Median
Endpoint
Sac/Val
N=1239
Valsartan
N=1256
Rate Ratio
(95% CI) Rate Ratio (95% CI)
Primary composite endpoint 457 591 0.78 (0.64 –0.95)
Total HF hospitalizations
332 463 0.75 (0.60 –0.95)
CV death
125 128 0.99 (0.77 – 1.26)
Expanded composite endpoint 480 622 0.78 (0.64 –0.94)​
Total Worsening HF* 355 494 0.74 (0.60 –0.93)
0.50.60.70.80.911.11.3
Val BetterSac/Val Better

74. 74
•Sacubitril/Valsartan safety profile well-characterized
‒Over 2.6 million patient-years of post-marketing safety data in HFrEF
•Clinical development program with over 23,000 CHF patients
•Label adequately addresses potential risks
Safety
Existing Favorable Safety Profile in HFrEF

75. 75
Summary of Adverse Events During Double-Blind Period
Sacubitril/Valsartan
N=2419
%
Valsartan
N=2402
%
Patients with at least one AE 95.1 95.5
Patients with at least one Serious AE (SAEs) 58.9 59.0
Patients who died 14.3 14.9
Patients who discontinued due to AEs 20.4 21.7
Discontinued due to SAEs 13.5 14.1
Discontinued due to non- serious AEs 7.4 8.4
AEs, SAEs, Deaths, and Discontinuation due to AEs were Similar in the Two Groups
N: Total number of patients in the treatment group in this safety analysis set.

76. 76
Safety Topics of Interest
Hypotension Occurred More Frequently with Sacubitril/Valsartan
Renal Impairment and Hyperkalemia Occurred More Frequently with Valsartan
Angioedema was Rare but Numerically Greater in Sacubitril/Valsartan. No Case Involved Airway Compromise.
Sacubitril/Valsartan
N = 2419
(%)
Valsartan
N = 2402
(%)
Hypotension SBP <100 mmHg 15.7 10.7
Hypotension AEs 23.2 17.0
Hypotension AEs (SMQ) 33.2 26.9
Hyperkalemia Serum potassium ≥5.5 mmol/L 17.8 19.6
Serum potassium >6.0 mmol/L 3.1 4.2
Hyperkalemia AEs (NMQ) 11.2 15.1
Renal impairmentSerum creatinine >2.0 mg/dL 10.9 13.8
Serum creatinine >2.5 mg/dL 3.8 4.4
Serum creatinine >3.0 mg/dL 1.6 1.7
Renalimpairment AEs (SMQ) 20.3 23.7
Angioedema* 0.6 0.2
N: Total number of patients in the treatment group in this safety analysis set; NMQ: Novartis MedDRA Query, SMQ: StandardizedMedDRA Query.
* Adjudication confirmed by AAC

77. 77
Combined PARADIGM -HF &
PARAGON-HF Results

78. 78
<<McMurray et al., Eur J Heart Fail 2013;15:1062– 73>>
<<Solomon et al. J Am Coll Cardiol HF 2017;5:471– 82>>
Pre-specified Pooled Analysis Across PARAGON-HF and
PARADIGM-HF with Similar Entry Criteria
PARADIGM-HF
•Age ≥ 18 years
•Signs/symptoms of heart failure (NYHA II- IV)
•LVEF ≤ 40%
•Elevation NT-proBNP (400/600 or BNP 100/150
depending on prior HF hospitalization)
PARAGON-HF
•Age ≥ 50 years
•Signs/symptoms of heart failure (NYHA II- IV)
•LVEF≥ 45%
•Elevation NT-proBNP (200/600 or 300/900
depending on prior HF hospitalization and AF)
•Structural heart disease (LVH or LAE)
•No prior LVEF < 40%
Entry Criteria Study Design
Randomization
N=4822 Double-blind, randomized treatment period
Sacubitril/Valsartan 97/103 mg BID
Valsartan 160 mg BID
On top of optimal background medications for comorbidities
(excluding ACEI and ARB)
Event-driven, median follow-up 35 months~2 weeks 1-4 weeks 2-4 weeks
Screening
Valsartan
80 mg BID
Sac/Val
49/51 mg BID
Single-blind run-in period
Sacubitril/Valsartan 97/103 mg BID
Enalapril 10 mg BID
On top of standard HF therapy (excluding ACEi and ARB)
Event-driven, median follow-up 27 months
Double-blind, randomized treatment period
Enalapril
10 mg BID
Sac/Val
49/51 mg BID
Sac/Val
97/103 mg BID
Single-blind run-in period
2 weeks 1-2 weeks 2-4 weeks
Randomization
N=8442

79. 79
Sacubitril/Valsartan Compared to RAS inhibitor
Pre-specifiedPooling of PARAGON-HF and PARADIGM-HF (N=13,195)
0.003
P-value
< 0.001
< 0.001
<<Solomon et al. Circulation. 2020;141:352– 361>>
<<Solomon et al. AHA 2019>>
(hazard ratio)
(rate ratio)
(rate ratio)
P-value
(hazard ratio)
<0.001
<0.001
<0.001
(hazard ratio)
(hazard ratio)

80. 80
Total HF Hospitalizations and CV Death by LVEF Septile Across
Spectrum of Ejection Fraction (PARAGON- HF and PARADIGM-HF)
LVEF Septile Entresto
EAR
Comparator
EAR
Rate Ratio
Estimate (95% CI)
Rate Ratio
Estimate (95% CI)
≤ 25% 18.8 23.0 0.816(0.684,0.973)
>25% to ≤ 30% 14.9 20.7 0.720(0.590,0.878)
>30% to ≤ 33% 14.6 18.9 0.767(0.575,1.023)
>33% to ≤ 36% 12.9 16.7 0.759(0.589,0.978)
>36% to ≤ 50% 13.1 15.1 0.878(0.708,1.087)
>50% to ≤ 60% 12.4 15.4 0.794(0.632,0.996)
> 60% 13.3 11.8 1.117(0.855,1.458)
FavorsSac/Val Favors Comparator
EAR: exposure- adjusted rate
0.4 1.610.6 0.8 1.21.4

81. 81
Treatment Effect Across Left Ventricular Ejection
Fraction
Sac/Val better
RASi better
Rate Ratio (Sacubitril/Valsartan vs.
RASi)
Ejection Fraction (%)
PARAGON-HF and PARADIGM-HF
<<Solomon et al. Circulation. 2020;141:352– 361>>
15 20 25 30 35 40 45 50 55 60 65 70 75
1.4
1.2
1
0.8
0.6
Total HF Hospitalizations and CV Death
RASi: renin- angiotensin- aldosterone– system inhibitor

82. 82
Treatment Effect by LVEF and Sex
Ejection Fraction (%)
Benefit Extends to Higher LVEF in Women (PARAGON-HF and PARADIGM-HF)
Total HF Hospitalizations and CV Death
Rate Ratio (Sacubitril/Valsartan vs.
RASi)
2
1
0.8
0.6
0.5
15 2025 303540 45505560657075
Men
Women
<<Solomon et al. Circulation. 2020;141:352– 361>>
RASi: renin- angiotensin- aldosterone– system inhibitor

83. 83
•Building on the pharmacodynamic and morphological benefits demonstrated in the Phase II trial,
PARAGON-HFwas the largestand only active-controlled PhaseIII studyinHFpEF
•Despite narrowly missing statistical significance for primary endpoint, the totality of evidence support a real
albeit modest benefit in the overall study population:
–Supportive analyses of the primary endpoint show a consistent treatment effect (13%-16%)
•Expanded composite endpoint incorporating urgent HF visits
•Investigator-reported endpoint events
•FDA-recommended re-adjudication of CEC -unconfirmed investigator reported HF hospitalizations
–Secondary efficacy endpoints (NYHA, KCCQ and renal composite endpoint) consistently favor
sacubitril/valsartan
•No new safety signals and the overall safety profile in HFpEF similar to HFrEF
•Benefit observed in overall trial population driven by large reductions in the total worsening HF events in
HFpEF patients with LVEF below normal
Summary of Sacubitril/Valsartan Effects in HFpEF

84. 84
PARAGON-HF:
Clinical implications?
John McMurray, MD
BHF Cardiovascular Research Centre,
University of Glasgow & Queen Elizabeth University Hospital in Glasgow

85. 85
Starting point: we have no approved therapies for HFpEF –
unmet need++
•Is sacubitril/valsartan beneficial in HFpEF?
•
Is sacubitril/valsartan beneficial in allpatients with HFpEF?
•If sacubitril/valsartan is beneficial in HFpEF, is the benefit
worthwhile?
•Is the safety profile of sacubitril/valsartan in HFpEF
acceptable?
Outline

86. 86
PARAGON- HF Primary Composite Endpoint Results
*Semiparametric LWYY method
Recurrent Event Analysis of Total HF Hospitalizations and CV Death*
Valsartan (n = 2389)
1009 events, 14.6 per 100 pt-years
Sacubitril/Valsartan (n = 2407) 894 events, 12.8 per 100 pt-years
Rate ratio 0.87 (95% CI 0.75, 1.01)
P= 0.059
Total HF hospitalizations and CV death
Mean cumulative events per 100 patients
Years
55
0
5
10
15
20
25
30
35
40
45
50
0 1 2 3 4

87. 87
Summary of Primary and Expanded Composite
Endpoint Results
Endpoint Sac/Val
N=2407
Valsartan
N=2389
Rate Ratio
(95% CI)
Rate Ratio (95% CI) 2-sided
P-value
CEC-confirmed
Primary endpoint 894 1009 0.870 (0.753, 1.005) 0.059
Expanded composite endpoint*934 10640.861 (0.747, 0.993) 0.040
Supportive analyses of the primary endpoint
Investigator reported
primary endpoint events
1064 12410.843 (0.736, 0.966) 0.014
Investigator reported expanded
composite endpoint**
1200 14140.834 (0.733, 0.950) 0.006
Re-adjudication analysis
of primary endpoint events
999 11350.865 (0.751, 0.995) 0.043
* Exploratory endpoint including CV death and total HF hospitalizations + urgent HF visits
** Investigator reported CV death and total HF hospitalizations + urgent HF visits
0.4 10.6 0.8 1.21.4
sacubitril/valsartan bettervalsartan better

88. 88
Secondary Endpoint Results
KCCQ, NYHA and Renal Endpoints C onsistently Favor Sacubitril/Valsartan
HR:hazard ratio; LSM:least square mean; OR:odds ratio
*defined as renal death, reaching end stage renal disease (ESRD), or ≥50% decline in estimated glomerular filtration rate (eGFR)relative to baseline
A sequential multiple testing procedure was used for the secondary analyses in the order of 1) KCCQ and NYHA 2) composite renal endpoint. All-cause mortality was not included in the multiple testing
strategy and was planned to be tested at full alpha level after rejection of the primary hypothesis.
Sacubitril/
Valsartan
Valsartan
Effect Size
(95% CI)
KCCQ clinical summary score at 8 months – n
LSM of changefrom baseline(SE)
2250
-1.5 (0.4)
2226
-2.5(0.4)
LSM of difference
1.03 (-0.00 –2.06)
NYHAfunctional classification at 8 months –n
Change from baseline -%
Improved
Unchanged
Worsened
2316
15.0%
76.3%
8.7%
2302
12.6%
77.9%
9.6%
OR for improvement
1.45 (1.13 –1.86)
Composite renal endpoint* - n/N
33/2407
1.4%
64/2389
2.7%
HR = 0.504
(0.331 –0.767)
All-cause mortality –n/N
342/2407
14.2%
349/2389
14.6%
HR = 0.970
(0.835 –1.126)

89. 89
Endpoint
RR/HR/OR (95% CI)
Sac/Val versus Val
RR/HR/OR (95% CI)
Sac/Val versus ValZ Score*
CEC confirmed: CV death and total HFH RR 0.87 (0.75, 1.01) -1.84
Change in NYHA class from baseline to 8 months OR 0.69 (0.54, 0.89) -2.63
Renal composite outcome HR 0.50 (0.33, 0.77) -3.20
Death from any cause HR 0.97 (0.84, 1.13) -0.41
Diff (Val –Sac/Val)
Change in KCCQ clinical summary score at 8 months -1.03 (-2.06, 0.00) -1.93
Average z score -2.00
p-value (2-sided) 0.0008
Effects of Sacubitril/Valsartan on the Primary and
Secondary Endpoints: Just due to chance?
0.4 0.60.811.21.62 3
-3 -2 -1 0 1 2 3
Favors ValsartanFavors Sac/Val
For NYHA, the treatment effect was expressed in terms of the odds for unfavorable NYHA class changes so that favorable changes f or sacubitril/valsartan
appear on the left side of the figure (note that 0.69=1/1.45)
* Observed z scores under simplified models; negative values indicate treatment effect in favor of Sacubitril/Valsartan.
The treatment effect distribution under the null hypothesis was simulated based on 100,000 random treatment permutations (withinregions).

90. 90
Starting point: we have no approved therapies for HFpEF –
unmet need++
•Is sacubitril/valsartan beneficial in HFpEF?
•
Is sacubitril/valsartan beneficial in allpatients with HFpEF?
•If sacubitril/valsartan is beneficial in HFpEF, is the benefit
worthwhile?
•Is the safety profile of sacubitril/valsartan in HFpEF
acceptable?
Outline

91. 91
PARAGON-HF: Pre -Specified Subgroup Analyses
(Primary Outcome)
Left ventricular ejection fraction
Sex
<<Solomon SD et al. N Engl J Med 2019;381:1609-1620.>>

92. 92
•Subgroup must be prespecified (not post hoc)
•Should be large –patients and events (smaller less reliable –subgroups
are always underpowered)
•Use a test for interaction and ideally adjust for multiplicity (ideally a
multivariable analysis)
•Examine the architecture of the total data and adjacent subgroups
(internal consistency)
•Interpret the results in the context of similar data from other trials (external
validation)
•Biological coherence/plausibility
How Do We Interpret Subgroups?

93. 93
•Subgroup must be prespecified (not post hoc) 
•Should be large –patients and events (smaller less reliable –subgroups
are always underpowered) 
•Use a test for interaction and ideally adjust for multiplicity (ideally a
multivariable analysis) 
•Examine the architecture of the total data and adjacent subgroups
(internal consistency)
•Interpret the results in the context of similar data from other trials (external
validation)
•Biological coherence/plausibility
How Do We Interpret Subgroups?

94. 94
Subgroup
No. of
Events/Patients
Rate Ratio
(95% CI)
Overall EF 1903/4796 0.87 (0.75– 1.01)
≤50 512/1208 0.82 (0.63– 1.06)
>50–57 536/1287 0.77 (0.57– 1.03)
>57–63 467/1202 0.91 (0.68– 1.22)
>63 388/1099 1.09 (0.80– 1.47)
Treatment Effect by Ejection Fraction Quartiles
Primary Composite Total HF Hospitalizations and CV Death
Subgroup estimates are based on re- running the primary analysis model for each of the four subgroups defined by the by LVEF quartiles.
0.4 1.610.6 0.8 1.21.4

95. 95
Total HF Hospitalizations and CV Death by LVEF Septile Across
Spectrum of Ejection Fraction (PARAGON- HF and PARADIGM-HF)
LVEF Septile Entresto
EAR
Comparator
EAR
Rate Ratio
Estimate (95% CI)
Rate Ratio
Estimate (95% CI)
≤ 25% 18.8 23.0 0.816(0.684,0.973)
>25% to ≤ 30% 14.9 20.7 0.720(0.590,0.878)
>30% to ≤ 33% 14.6 18.9 0.767(0.575,1.023)
>33% to ≤ 36% 12.9 16.7 0.759(0.589,0.978)
>36% to ≤ 50% 13.1 15.1 0.878(0.708,1.087)
>50% to ≤ 60% 12.4 15.4 0.794(0.632,0.996)
> 60% 13.3 11.8 1.117(0.855,1.458)
FavorsSac/Val Favors Comparator
EAR: exposure- adjusted rate
0.4 1.610.6 0.8 1.21.4

96. 96
How Do We Interpret Subgroups?
•Subgroup must be prespecified (not post hoc) 
•Should be large – patients and events (smaller less reliable –
subgroups are always underpowered) 
•Use a test for interaction and ideally adjust for multiplicity (ideally a
multivariable analysis) 
•Examine the architecture of the total data and adjacent subgroups
(internal consistency) 
•Interpret the results in the context of similar data from other trials
(external validation)
•Biological coherence/plausibility

97. 97
•Subgroup must be prespecified (not post hoc) 
•Should be large – patients and events (smaller less reliable –
subgroups are always underpowered) 
•Use a test for interaction and ideally adjust for multiplicity (ideally a
multivariable analysis) 
•Examine the architecture of the total data and adjacent subgroups
(internal consistency) 
•Interpret the results in the context of similar data from other trials
(external validation)
•Biological coherence/plausibility
How Do We Interpret Subgroups?

98. 98
Effects of Candesartan and Spironolactone
According to LVEF
Because of the growing
interest in HFmrEF
(“mid-range” or
“mildly-reduced” EF)

99. 99
CV death or HF hospitalizations (time-to -first)
Candesartan and MRAs: TreatmentEffect by LVEF
Adapted from Dewan P et al Eur J Heart Fail 2020;22: 898-901

100. 100
CV death or HF hospitalizations (time-to -first)
TreatmentEffect by LVEF: 3 Different Therapies
<<Dewan P et al Eur J Heart Fail 2020;22: 898-901>>

101. 101
25 30 35 40 45 50 55 60
LVEF (%)
Systolic Dysfunction and Favourable Response to
Treatment Over Wider Range of LVEF?
HFmrEF HFpEF
HFrEF
Candesartan
Spironolactone
Sacubitril/valsartan

102. 102
CV death or HF hospitalizations (time-to -first)
TreatmentEffect by LVEF and Sex
Benefit Extends to Higher LVEF in Women
<<Dewan et al. Eur J Heart Fail 2020; 22:898- 901>>

103. 103
Starting point: we have no approved therapies for HFpEF –
unmet need++
•Is sacubitril/valsartan beneficial in HFpEF?
•Is sacubitril/valsartan beneficial in all patients with HFpEF?
•If sacubitril/valsartan is beneficial in HFpEF, is the benefit
worthwhile?
•Is the safety profile of sacubitril/valsartan in HFpEF
acceptable?
Outline

104. 104
Estimated Absolute Benefit: Potential Events
Prevented by Treating 1000 Patients for 3 Years
*Calculations are based on the between treatment group differences in exposure- adjusted event rates
Primary
composite events
Prevented*
HF
hospitalizations
prevented*
PARADIGM-HF 122 76
PARAGON-HF: Overall 54 49
PARAGON-HF: LVEF ≤median 108 106
If there really is an effect in patients with a LVEF≤57%, it may be substantial

105. 105
Starting point: we have no approved therapies for HFpEF –
unmet need++
•Is sacubitril/valsartan beneficial in HFpEF?
•Is sacubitril/valsartan beneficial in all patients with HFpEF?
•If sacubitril/valsartan is beneficial in HFpEF, is the benefit
worthwhile?
•Is the safety profile of sacubitril/valsartan in HFpEF
acceptable?
Outline

106. 106
•Safety profile of sacubitril/valsartan confirmed
•PARAGON-HF demonstrated:
‒Consistent safety profile with PARADIGM-HF
‒Similar safety profile to valsartan
‒No new safety signals
•Hypotension and angioedema are currently addressed in the
sacubitril/valsartan label
•Sacubitril/valsartan results in lower rates of renal dysfunction and
hyperkalemia compared to valsartan
Sacubitril/Valsartan in HFpEF: Summary of Safety
Findings

107. 107
•HFpEFis a syndrome (or syndromes –HFmrEF too?), causing disabling
symptoms and frequent hospitalization in older individuals, many of whom are
women, and for which there is no approved treatment
•The totality of evidence from PARAGON-HF supports the conclusion that
sacubitril/valsartan has clinical benefits, and a favorable benefit-to-risk profile, in
HFpEF, overall
•In a prespecified subgroup analysis, sacubitril/valsartan seemed to have more
benefit in patients with a LVEF at or below the median of 57% (and in women)
•Additional data to show that other drugs working on neurohumoral pathways also
seem to be beneficial in patients with a LVEF in lower part of the HFpEF range
•The potential reduction in events with sacubitril/valsartan in patients with a LVEF
≤median in PARAGON- HF was substantial and the safety profile is acceptable
PARAGON- HF: Summary

108. 108
Closing Remarks
David Soergel, MD
Development Unit Head, Cardiovascular, Renal & Metabolism
Novartis Pharmaceuticals Corporation

109. 109
Totality of Evidence Supports Benefit in HFpEF
•HFpEF is a debilitating disease with no approved therapy
•Entresto is approved in the HFrEF population
•Evidence supports extending use to adjacent HFpEF population to
reduce worsening HF events
‒Greater benefit in patients with LVEF below normal
•Favorable safety profile

110. 110
Sponsor Backup Slides Shown
Cardiovascular and Renal Drugs Advisory Committee
December 15, 2020

111. 111
Country Cases Country CasesCountry Cases Country Cases
Germany 52 Singapore 7 Bulgaria 5 Poland 2
United States 47 Brazil 7 South Africa 5 Switzerland 2
China 25 Croatia 7 Israel 4 Turkey 1
Italy 20 Guatemala 7 Austria 3 Philippines 1
Canada 16 Japan 7 United Kingdom 3 South Korea 1
Spain 15 Hungary 6 France 3 Colombia 1
Belgium 13 Taiwan 5 Czech Republic 3 Russian Federation 1
Australia 12 Netherlands 5 Denmark 3 Slovenia 1
Slovakia 12 Sweden 5 Argentina 2 Mexico 1
ALL008- 13
Geographic Distribution of Investigator- reported
Urgent HF Visits (N = 310)

112. 112
Trial/cohort Event type Number needed to treat (95% CI)
PARADIGM-HF Primarycomposite endpoint 8 (6 –12)
PARADIGM-HF HF hospitalization 13 (10 – 21)
PARAGON-HF LVEF ≤median Primary composite endpoint 9 (6 –18)
PARAGON-HF LVEF ≤median HF hospitalization 9 (7 –17)
EFF101-14
Number Needed to Treat for 36 Months to
Prevent One Event
PARAGON-HF low LVEF vs. PARADIGM-HF
Primary composite endpoint based on relative rate reduction of total HF hospitalizations and CV death

113. 113
•Sacubitril/Valsartan significantly reduced the risk of the composite renal endpoint, primarily driven by
lower incidence of ≥50% reduction in eGFR
Composite Renal Endpoint
* indicates nominal statistical significance (2- sided) with an alpha level of 0.05
A hazard ratio <1 indicates an effect in favor of Sac/Val.
Composite
Sacubitril/Valsartan
N=2407
n (%)
Valsartan
N=2389
N (%)
Odds Ratio
(95% CI)
P-value
(2-sided)
Composite renal endpoint
33/2407
(1.4)
64/2389
(2.7)
0.504
(0.331 –0.767)
0.001*
(i) Renaldeath
1/2407
(<0.1)
1/2389
(<0.1)
0.930
(0.058 –14.861)
0.959
(ii) Reaching end stage renal disease
(ESRD)
7/2407
(0.3)
12/2389
(0.5)
0.577
(0.227 –1.467)
0.248
(iii)≥50% decline in estimated glomerular
filtration rate (eGFR) relative to baseline
27/2407
(1.1)
60/2389
(2.5)
0.441
(0.280 –0.694)
<0.001*
•Similar effect seen in PARADIGM-HF with lower risk of the same renal composite endpoint with
sacubitril/valsartan compared with enalapril (HR = 0.678; 95% CI 0.462 –0.996; p = 0.0475)
ALL014- 2

114. 114
No. of Events Patients* (n)
1 event 671
2events 234
3 events 88
4 events 32
5 events 24
6 events 20
≥ 7 events 14
Total 1083
•First events: 1083/1903 = 57%
•Recurrent events: 820/1903 = 43%
EFF201-3
* Patients with specific number of events
Number of Patients with 1 or More Primary Events
38% (412/1083) of patients had recurrent events

115. 115
Descriptive Statistics of Contribution of K
th
Event to Treatment
Difference in Number of Composite Primary Endpoint Events in
PARAGON- HF
Variable
Sac/Val
N = 2407
Valsartan
N = 2389
k
th
event contribution to
treatment difference
Total no. of primary composite endpoints 894 1009 115 (100%)
No. of patients with k
th
event, n
1
st
event 526 557 31 (26.96)
2
nd
event 192 220 28 (24.35)
3
rd
event 84 94 10 (8.70)
4
th
event 41 49 8 (6.96)
5
th
event 25 33 8 (6.96)
6
th
event 15 19 4 (3.48)
7
th
event 4 10 6 (5.22)
8
th
event 1 8 7 (6.09)
9
th
event 1 5 4 (3.48)
10
th
event 1 4 3 (2.61)
>10
th
event 1 3 6 (5.22)
EFF201-2

116. 116
Truncated LWYY Analysis of the Primary Composite
Endpoint by Censoring after the K
th
Event (FAS)
•Relative contribution of 8th and subsequent events to the overall treatment effect was approximately 13%
EFF201-4
K No. of Events Rate Ratio Rate Ratio
Sac/Val Valsartan Estimate (95% CI) Estimate (95% CI)
4 843 920 0.898(0.789, 1.023)
5 868 953 0.892(0.781, 1.020)
6 883 972 0.890(0.776, 1.020)
7 887 982 0.885(0.771, 1.017)
8 888 990 0.880(0.765,1.011)
No truncation 894 1,009 0.870(0.753,1.005)
Favors Sac/Val Favors Valsartan
0.7 0.8 0.9 1 1.11.21.3