Valsartan Initiation and Health Status Outcomes in Heart Failure With Reduced Ejection Fraction

JACC: HEART FAILURE

VOL.

-, NO. -, 2019

PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

Association Between Sacubitril/Valsartan Initiation and Health Status Outcomes in Heart Failure With Reduced Ejection Fraction Yevgeniy Khariton, MD,a Gregg C. Fonarow, MD,b Suzanne V. Arnold, MD, MHA,a Ann Hellkamp, MS,c Michael E. Nassif, MD, MS,d Puza P. Sharma, MBBS,e Javed Butler, MD, MPH, MBA,f Laine Thomas, PHD,c Carol I. Duffy, DO,e Adam D. DeVore, MD, MHS,g Nancy M. Albert, PHD,h J. Herbert Patterson, PHARMD,i Fredonia B. Williams, EDD,j Kevin McCague, MA,e John A. Spertus, MD, MPHa

ABSTRACT OBJECTIVES This study sought to describe the short-term health status benefits of angiotensin-neprilysin inhibitor (ARNI) therapy in patients with heart failure and reduced ejection fraction (HFrEF). BACKGROUND Although therapy with sacubitril/valsartan, a neprilysin inhibitor, improved patients’ health status (compared with enalapril) at 8 months in the PARADIGM-HF (Prospective Comparison of ARNI with ACE inhibitor to Determine Impact on Global Mortality and Morbidity in Heart Failure) study, the early impact of ARNI on patients’ symptoms, functions, and quality of life is unknown. METHODS Health status was assessed by using the 12-item Kansas City Cardiomyopathy Questionnaire (KCCQ) in 3,918 outpatients with HFrEF and left ventricular ejection fraction #40% across 140 U.S. centers in the CHAMP-HF (Change the Management of Patients with Heart Failure) registry. ARNI therapy was initiated in 508 patients who were matched 1:2 to 1,016 patients who were not initiated on ARNI (no-ARNI), using a nonparsimonious time-dependent propensity score (6 sociodemographic factors, 23 clinical characteristics), prior KCCQ overall summary (KCCQ-OS) score and angiotensin-converting enzyme inhibitor/angiotensin receptor blocker status. RESULTS Multivariate linear regression demonstrated a greater mean improvement in KCCQ-OS in patients initiated on ARNI therapy (5.3
19 vs. 2.5
17.4, respectively; p < 0.001) over a median (interquartile range [IQR]) of 57 (32, 104) days. The proportions of ARNI versus no-ARNI groups with $10-point (large) and $20-point (very large) improvements in KCCQ-OS were 32.7% versus 26.9%, respectively, and 20.5% versus 12.1%, respectively, consistent with numbers needed to treat of 18 and 12, respectively. CONCLUSIONS In routine clinical care, ARNI therapy was associated with early improvements in health status, with 20% experiencing a very large health status benefit compared with 12% who were not started on ARNI therapy. These findings support the use of ARNI to improve patients’ symptoms, functions, and quality of life. (J Am Coll Cardiol HF 2019;-:-–-) Published by Elsevier on behalf of the American College of Cardiology Foundation.

From the aDepartments of Cardiology and Cardiovascular Outcomes Research, Saint Luke’s Mid America Heart Institute/University of Missouri-Kansas City, Kansas City, Missouri; bDepartment of Cardiology, Ahmanson-UCLA Cardiomyopathy Center, Ronald Reagan University of California Los Angeles Medical Center, Los Angeles, California; cDuke Clinical Research Institute, Durham, North Carolina; dDepartment of Cardiology, Washington University School of Medicine in Saint Louis, Saint Louis, Missouri; eNovartis Pharmaceuticals Corp, East Hanover, New Jersey; fDepartment of Medicine, University of Mississippi Medical Center, Jackson, Mississippi; gDivision of Cardiology, Department of Medicine, and the Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; hCleveland Clinic, Cleveland, Ohio; iEshelman School of Pharmacy, University of North Carolina, Chapel Hill, North Carolina; and the jMended Hearts, Huntsville, Alabama. The CHAMP-HF (Change the Management of Patients with Heart Failure) and the present study were funded by the Novartis Pharmaceuticals Corp. Drs. Khariton and Nassif are supported by U.S. National Heart, Lung, and Blood Institutes award T32HL110837. The content is solely the responsibility of the authors and does not necessarily represent the official views of the U.S. National Institutes of Health (NIH). Dr. Khariton had full access to all the data in the study and takes responsibility for the integrity of the data and the

ISSN 2213-1779/$36.00

https://doi.org/10.1016/j.jchf.2019.05.016

2

Khariton et al.

JACC: HEART FAILURE VOL.

ABBREVIATIONS

A

-, NO. -, 2019 - 2019:-–-

Sacubitril/Valsartan and Health Status Outcomes

primary treatment goal in patients

patient-reported health status. The CHAMP-HF study

with heart failure and reduced ejec-

was a prospective, multicenter, observational registry

tion fraction (HFrEF) is to optimize

of outpatients with HFrEF that captured serial health

their health status (i.e., symptoms, func-

status outcomes by using the 12-item Kansas City

tions, and quality of life) (1). Health status

Cardiomyopathy Questionnaire (KCCQ) (15), making it

reduced ejection fraction

is not only important from patients’ and pro-

an ideal data source with which to describe the health

KCCQ = Kansas City

viders’ perspectives (2) but also a strong and

status benefits of ARNI in routine clinical practice.

Cardiomyopathy Questionnaire

independent

AND ACRONYMS ARNI = angiotensin-neprilysin inhibitor

HFrEF = heart failure with

predictor

of

cardiovascular

morbidity and mortality (3–6). Accordingly, regulato-

METHODS

ry agencies are increasingly recognizing the significance of systematically quantifying patients’ health

STUDY DESIGN. CHAMP-HF is a multicenter, obser-

status by using patient-reported outcomes, to assess

vational registry designed to capture the care and

the health status benefits of novel therapies (7–9).

outcomes of patients with HFrEF across heteroge-

However, despite this era of rapidly expanding treat-

neous outpatient practices in the United States (14).

ments, few pharmacological interventions in HFrEF

Eligibility criteria included a diagnosis of chronic

have been shown to improve patients’ quality of life and reduce their symptom burden (1,10).

HFrEF with left ventricular ejection fraction #40% and use of $1 oral pharmacotherapy management of

The PARADIGM-HF (Prospective Comparison of

heart failure. Minimum required exposure to ARNI

ARNI with ACE inhibitor to Determine Impact on

was defined as uninterrupted ARNI therapy at any

Global Mortality and Morbidity in Heart Failure) trial

dose, either at the time of or between the most recent

compared sacubitril/valsartan therapy (angiotensin

KCCQ assessment and the next closest assessment

receptor-neprilysin inhibitor [ARNI]) with that of

occurring at least 2 weeks after the initiation of ARNI

enalapril and demonstrated improved survival and

treatment. Patients with limited life expectancy who

lower hospitalization rates with ARNI. The study also

were considered for advanced mechanical support

showed significantly less deterioration in patients’

(e.g., left ventricular assist device, heart trans-

health status with ARNI from baseline to 8 months (11).

plantation) and who required hemodialysis were

These data led the 2016 European and North American

excluded. Study coordinators at each practice site

guideline authors to recommend ARNI for patients

were responsible for consecutively identifying and

with HFrEF or as a replacement for angiotensin

enrolling patients during the course of a routine

converting-enzyme (ACE) inhibitor or angiotensin re-

outpatient HFrEF visit. For this analysis, other

ceptor blocker (ARB) in patients with HFrEF (12,13).

exclusion criteria included: 1) use of ARNI before

However, a limitation of the PARADIGM-HF trial was

enrollment;

2)

documented

contraindication

or

that patients’ health status was not assessed before the

intolerance to ARNI; and 3) completion of <2 KCCQ

run-in phase, precluding an assessment of the early

assessments (1 before or at the initiation of ARNI and

health status benefits of ARNI. Moreover, the effec-

1 at least 2 weeks after initiation) (Figure 1). The

tiveness of ARNI in patients’ health status in routine

CHAMP-HF study was funded by Novartis. All

clinical practice is unknown.

participating sites received institutional review board

To address these gaps in knowledge, this study

approval, and informed consent was signed by each

used data from the CHAMP-HF (Change the Manage-

participant before enrollment.

ment of Patients with Heart Failure) registry (14) to

DATA COLLECTION AND PRIMARY ANALYSIS OUTCOME.

examine the association between ARNI therapy and

At

the

time

patients

were

enrolled,

accuracy of the data analysis. Dr. Sharma was an employee of Novartis at the time of the development and review of this paper. Dr. Spertus has received research support from National Institute of Health (NIH), American College of Cardiology Foundation, Bayer, Novartis, and Abbott Vascular; and he serves on a Scientific Advisory Board for United Healthcare; and is a consultant for Novartis, V-Wave, AstraZeneca, Janssen, Corvia, and Bayer; and holds patent rights to the Kansas City Cardiomyopathy Questionnaire; and he holds equity in Health Outcomes Sciences. Dr. Thomas has received research support from Novartis Pharmaceuticals Corp. Dr. Fonarow has received research support from NIH; and is a consultant for Abbott, Amgen, Bayer, Janssen, Medtronic, and Novartis. Dr. DeVore has received research support from the American Heart Association, Amgen, NIH, and Novartis; and is a consultant for Novartis. Dr. Butler has received research support from NIH and European Union; and is a consultant for Amgen, Bayer, Boehringer Ingelheim, Cardiocell, CVRx, Gilead, Janssen, Medtronic, Merck, Novartis, Relypsa, and ZS Pharma. Dr. Albert is a consultant for Novartis, AstraZeneca, and Boston Scientific; and has received honoraria from Novartis. Dr. J. Herbert Patterson reports consulting and research support for Novartis. Drs. Duffy and McCague are employees of Novartis. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received April 26, 2019; revised manuscript received May 28, 2019, accepted May 29, 2019.

study

JACC: HEART FAILURE VOL.

-, NO. -, 2019

- 2019:-–-

Khariton et al. Sacubitril/Valsartan and Health Status Outcomes

F I G U R E 1 Patient Exclusion Flowsheet

Patients receiving ARNI initiation before enrollment, contraindications to ARNI treatment, and missing covariate and KCCQ data were excluded. ACE ¼ angiotensin-converting enzyme; ARB ¼ angiotensin receptor blocker; ARNI ¼ angiotensin-neprilysin inhibitor; CHAMP-HF ¼ Change the Management of Patients with Heart Failure; KCCQ ¼ Kansas City Cardiomyopathy Questionnaire.

coordinators interviewed them to collect their base-

specific patient-reported outcome instrument that

line sociodemographic and health status information

measured patients’ health status over the preceding

and performed chart abstraction to establish a

2 weeks that preserves the psychometric properties

comprehensive clinical history and medication pro-

of the KCCQ 23-question version used in the

file. On all subsequent visits (1, 3, 6, and 12 months),

PARADIGM-HF trial (15). The KCCQ overall summary

patient-reported data were collected either during in-

score (KCCQ-OS) consists of 4 equally-weighted do-

person or telephone interviews. The study did not

mains: physical limitation (KCCQ-PL), symptom fre-

dictate or recommend any changes in therapy nor

quency (KCCQ-SF), quality of life (KCCQ-QoL), and

mandatory laboratory measurements.

social limitation (KCCQ-SL), which were secondary

The primary outcome for this analysis was changes

outcomes of the study. All domains and the KCCQ-OS

in scores of the KCCQ, a well-validated disease-

scores range from 0 to 100, where higher scores

3

4

Khariton et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Sacubitril/Valsartan and Health Status Outcomes

indicate better health status (15). Prior and extensive

and their most recent KCCQ scores. Due to non-

work has defined the clinical significance of both

mandatory reporting of laboratory data, these char-

group mean and individual patient-level changes.

acteristics were not included in the model due to high

Patient-level changes of <5, 5 to <10, 10 to <20,

missing rates. Left ventricular ejection fraction,

and $20 points represent worse to small, moderate,

health status data, and medications were updated

large, and very large improvements, respectively

throughout the analysis, whereas physiological data

(3,6,16,17).

(e.g., vital signs) were updated every 6 months (the

STUDY COHORTS AND DEFINING ARNI USE. Partici-

pants were allocated to receive ARNI therapy or not (no-ARNI treatment group), contingent on whether they began ARNI at any time after enrollment. Because ARNI might have been a preferred initial treatment, or patients might have switched from an ACE inhibitor/ARB to an ARNI, patients were matched directly on their pre-ARNI ACE inhibitor/ARB status (use of an ACE inhibitor/ARB within the preceding 2 weeks) (Online Figure 1). Due to the time-dependent nature

of

this

analysis,

if

a

no-ARNI

patient

completed $2 KCCQ assessments after initial propensity matching and before ARNI initiation, that patient was then potentially eligible to be included twice in the analysis, initially as a no-ARNI patient using the pre-ARNI KCCQ data and then as an ARNI patient using the post-ARNI KCCQ data.

first value within each 6-month period was used for the matching within that period). Continuous variables were assessed for linearity of their relationship with the primary outcome of the propensity analysis by using restricted cubic splines. For categorical variables, multilevel variables (e.g., household income) were divided into binary categories (white vs. nonwhite, income <$25,000 vs. $$25,000, and fulltime vs. part-time employment vs. not working) to simplify interpretation. The quality of propensity matching was evaluated by using standardized differences: the absolute differences in means (or proportions) divided by the average SD. A standardized difference of <0.10 (10%) reflected good covariate balance between groups (18). Propensity models and matching were conducted separately for patients who were versus those who were not taking ACE inhibitors/ARBs, and then the

STATISTICAL ANALYSIS. Time-dependent propensity

cohorts were combined. To be included in the ACE

score matching. Because patients might have been

inhibitor/ARB cohort, patients were required to have

prescribed ARNI at different times throughout the

been treated with ACE inhibitor/ARB within 2 weeks

registry, matching was used to ensure that no-ARNI

before enrollment (otherwise, they were assigned to

patients were identified at the same time during

the no- ACE inhibitor/ARB cohort). These steps

follow-up as those newly prescribed ARNI. All eligible

ensured comparability of newly prescribed ARNI pa-

participants were propensity matched at an ARNI-to-

tients and switchers from ACE inhibitor/ARB to ARNI.

no-ARNI ratio of 1:2 based upon on a time-dependent

Medians (IQR) of time (days) from pre-match to

propensity score, their most recent KCCQ-OS and ACE

post-match KCCQ assessments were calculated. The

inhibitor/ARB status. Propensity scores were calcu-

primary analysis included all patients (ARNI vs.

lated using Cox proportional hazards models, where

no-ARNI), with comparisons of newly prescribed

“time to ARNI” (days from enrollment to ARNI initi-

(ARNI vs. no-ACE inhibitor/ARB) and switching

ation) was the dependent variable and all patient-

(ARNI vs. ACE inhibitor/ARB) reported as secondary

level predictors (except sociodemographic factors)

analyses.

were allowed to vary over time. Variables in the

Most patient-level variables had #5% missing data,

propensity score included 6 sociodemographic (age,

with the exception of household income (21%) and

sex, race, Hispanic ethnicity, household income, and

body mass index (8%). For covariates in the pro-

employment status) and 23 clinical characteristics

pensity model missing values were imputed using

including medical history (atrial fibrillation, ventricu-

single imputation with full conditional specification.

lar tachycardia/fibrillation, cardiac resynchronization

KCCQ scores were not imputed.

therapy, chronic obstructive pulmonary disease, cor-

Linear regression and responder analysis. Changes in

onary artery disease, diabetes mellitus, depression/

KCCQ scores were calculated as the difference be-

anxiety, essential hypertension, ischemic cardiomy-

tween post-match and pre-match KCCQ scores. After

opathy, current smoker, prior HF hospitalization,

comparing

chronic

(beta-

compared the mean differences in the matched co-

blocker, mineralocorticoid antagonist, loop diuretic

horts and further compared the association of ARNI

agent, hydralazine, digoxin, ivabradine); physiolog-

initiation with changes in health status using 5 linear

ical measurements (body mass index, systolic blood

regression models of change: KCCQ-OS as the primary

pressure, heart rate, left ventricular ejection fraction);

outcome and changes in KCCQ domain scores as

kidney

disease);

medication

use

the

propensity-matched

cohorts,

we

JACC: HEART FAILURE VOL.

-, NO. -, 2019

Khariton et al.

- 2019:-–-

5

Sacubitril/Valsartan and Health Status Outcomes

secondary outcomes. The regression models used generalized estimating equations to adjust for all

T A B L E 1 Patient Characteristics For Combined Matched Cohort

Post-Match

covariates and clustering of patients within practice sites. Because mean KCCQ scores represented a population average effect, the distribution of changes was

ARNI n ¼ 508

No-ARNI* n ¼ 1,016

Standardized Differences†

9.0

Sociodemographic Age, yrs

64.0 (12.9)

65.2 (12.7)

also described so that the proportion of patients with

Female

30 (154)

29 (291)

3.7

clinically important changes in health status could be

White

75 (381)

74 (750)

2.7

appreciated. For each KCCQ score, where there was a

Hispanic

9 (47)

12 (120)

8.3

significant effect in the main model, the proportion of

Household income <$25,000

34 (175)

37 (380)

6.2

patients (n ¼ %) across categories of KCCQ improve-

Employed Full- or Part-Time

26 (131)

22 (221)

9.5

ment (worse to small [<5 points], moderate [$5

Clinical Measurements

to <10 points], large [10 to <20 points], and very large

Body mass index

[$20 points] changes) were calculated (15). When statistically significant, the number needed to treat (NNT) was calculated by first finding the absolute risk reduction as [ARR ¼ P A

31.2 (7.3)

30.5 (7.5)

9.8

Systolic BP

120.3 (17.3)

118.8 (17.4)

8.9

Heart rate

74.2 (13.2)

74.4 (13.4)

1.8

LVEF (%)

28.8 (7.1)

29.5 (8.5)

9.2

Medical History Atrial fibrillation/flutter

38 (195)

40 (406)

3.2

represent the proportion of ARNI and no-ARNI pa-

History of VT/VF

23 (119)

23 (231)

1.6

tients, respectively, with a given level of health status

CRT therapy

10 (49)

9 (94)

1.3

COPD

27 (139)

29 (291)

2.8

Coronary artery disease

62 (314)

66 (672)

9.0

Diabetes

43 (219)

43 (439)

0.2

Depression or anxiety

34 (171)

36 (363)

4.3

Hypertension

83 (424)

83 (841)

1.8

Ischemic HF cause

41 (209)

42 (426)

1.6

value of #0.05 was considered statistically signifi-

Current smoker

20 (101)

20 (207)

1.2

cant. All analyses were performed using SAS version

Prior HF hospitalization

40 (205)

41 (419)

1.8

Chronic kidney disease

18 (92)

23 (232)

11.7



P N], where P A and PN

improvement and 95% confidence intervals (CIs) were calculated as ARR
1.96  SE, where SE is Wald SE. The reciprocals of these estimates were used to calculate the NNTs. All estimates were reported using 95% CI and a p

14.3 software (SAS Institute, Cary, North Carolina).

Medication

RESULTS

ACE inhibitor/ARB

53 (267)

53 (534)

0.0

Beta-blocker

97 (492)

94 (954)

14.1

MRA

49 (250)

42 (423)

15.3

STUDY COHORT. At the time of this analysis, 3,918

Loop diuretic agent

73 (373)

69 (699)

10.2

patients across 140 sites were enrolled in the CHAMP-

Hydralazine

4 (22)

5 (46)

1.0

HF study between 2015 and 2017, after excluding

Digoxin

17 (88)

14 (143)

8.9

those patients who had been prescribed ARNI before

Ivabradine

2 (8)

2 (17)

0.8

enrollment (n ¼ 608); or who were missing de-

Prescribed $2 HF medications

93 (474)

88 (899)

16.8

Prescribed $3 HF medications

67 (341)

57 (580)

20.8

Overall

64.1 (23.5)

63.3 (23.9)

3.3

SF

70.9 (24.5)

71.1 (24.5)

0.8

missing KCCQ data before or after ARNI initiation

QoL

58.8 (28.3)

59.3 (28.1)

1.9

(n ¼ 183) (Figure 1). Of these, 580 were newly pre-

SL

69.3 (27.9)

68.3 (29.0)

3.3

scribed ARNI and 508 (88%) were successfully

PL

67.6 (26.2)

65.1 (27.2)

9.6

mographic, medical history, or medication data (n ¼ 198); or who had documented contraindications or intolerance to ARNI (n ¼ 119); and those who had

matched with 1,016 patients who had not begun ARNI therapy at the same point during follow-up (Table 1, Online Figure 2). For those patients who began treatment with ARNI, 267 (53%) had been taking ACE inhibitor/ARB therapy and 241 (47%) had not, within 2 weeks of ARNI initiation. The matched groups were

KCCQ score

Values are % (n) or %. *Fifty-five ARNI patients were used as no-ARNI matches before their ARNI start date. †Standardized difference is calculated as the differences in means or proportions divided by the SD. A standardized difference >10% is considered unbalanced. ACE ¼ angiotensin-converting enzyme; ARB ¼ angiotensin receptor blocker; BP ¼ blood pressure; COPD ¼ chronic obstructive pulmonary syndrome; CRT ¼ cardiac resynchronization therapy; HF ¼ heart failure; KCCQ ¼ Kansas City Cardiomyopathy Questionnaire; LVEF ¼ left ventricular ejection fraction; MRA ¼ mineralocorticoid antagonists; PL ¼ physical limitation; QoL ¼ quality of life; SF ¼ symptom frequency; SL ¼ social limitation; VF ¼ ventricular fibrillation; VT ¼ ventricular tachycardia.

well balanced in sociodemographic and clinical characteristics, with the exception of chronic kidney disease (standardized difference: 11.7%), beta-blocker

and no-ACE inhibitor/ARB (241 ARNI and 482 no-ACE

(14.1%), mineralocorticoid antagonists (MRA) (15.3%),

inhibitor/ARB) cohorts (Online Tables 1 and 2, Online

and loop diuretic treatment (10.2%). Similar compa-

Figures 3 and 4).

rability between groups was observed within the ACE

Characteristics of the analytic cohort are presented

inhibitor/ARB (267 ARNI and 534 ACE inhibitor/ARB)

in Online Table 3. Of the total sample, 42.0% were 40

6

Khariton et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Sacubitril/Valsartan and Health Status Outcomes

T A B L E 2 Mean Change in Scores and Multivariate Regression-Adjusted Difference in Mean Score Change in Propensity-Matched Patients (Combined Model)

Multivariate Regression-Adjusted Difference in Mean Score Change

Mean Change in Scores Change in Score (post  pre)

p Value

p Value

Mean ARNI (n ¼ 508) (SD)

Mean no-ARNI (n ¼ 1,016)

Mean ARNI vs. No-ARNI Post-KCCQ Score

Differences in Mean Change Estimate (95% CI)

Covariate-Adjusted Difference in Mean Change Estimate (95% CI)

KCCQ overall summary

5.3
18.6

2.5
17.4

<0.001

3.2 (1.5–4.9)

2.9 (1.1–4.6)

0.001

KCCQ physical limitation*

4.8
24.8

2.0
22.2

0.006

3.2 (0.9–5.5)

3.3 (0.8–5.8)

0.009

KCCQ Score

ARNI vs. no-ARNI

KCCQ symptom frequency

4.3
20.8

1.5
21.2

0.004

2.9 (0.9–4.9)

2.6 (0.6–4.6)

0.010

KCCQ social limitation†

5.9
26.0

3.6
23.8

0.030

2.6 (0.3–5.0)

2.7 (0.3–5.1)

0.029

KCCQ-quality of life

6.4
23.9

2.7
24.1

<0.001

3.8 (1.6–6.0)

3.3 (1.1–5.4)

0.003

Values are mean
SD or mean (confidence interval [CI]). *Total of 116 sets (348 patients) were omitted for missing pre- or post-match physical limitation score for 1 or more members of the set. †Total of 76 sets (228 patients) were omitted for missing pre- or post-match social limitation score for 1 or more members of the set.

to 64 years of age and 44.4% were 65 to 80 years of

difference of 2.7 points [95% CI: 0.4 to 5.0; p ¼ 0.024])

age; and 29.2% were women, and 74.2% were white.

(Online Tables 4a, 4b, 5a, and 5b).

Cardiac and non-cardiac comorbidities were common,

RESPONDER ANALYSIS. The proportion of patients

with 64.6% having coronary artery disease, 27.9%

experiencing at least moderate, large, and very large

having chronic obstructive lung disease/asthma,

health status improvements according to changes in

33.7% having depression/anxiety, 42.7% having dia-

KCCQ-OS scores were calculated (Table 3). Overall,

betes, and 38.4% having atrial fibrillation/flutter.

43.7% of ARNI patients (vs. 39.8% of the no-ARNI

Most patients (62.0%) had no HF hospitalizations in

patients), 32.7% of ARNI patients (vs. 26.9% of no-

the year before enrollment, reflecting a relatively

ARNI patients), and 20.5% of ARNI patients (vs.

stable HF cohort. Mean systolic blood pressure and

12.1% of no-ARNI patients) experienced at least a

left ventricular ejection fraction in the cohort were

moderate ($5-point increase), large ($10-point in-

120
18 mm Hg and 28
8%, respectively. Use of HF

crease), and very large ($20-point increase) health

therapies in the matched cohorts was high, including

status benefit, respectively (Online Figures 5 and 6).

beta-blockers in 94.3%, MRAs in 41.4%, and loop

Based on these results, it was found that for every 18

diuretic agents in 69.5%. A minority were prescribed

patients (95% CI: 10 to 111) and 12 patients (95% CI: 9

digoxin (14.6%) and ivabradine (1.6%). The average

to 24) started on ARNI therapy, 1 more patient would

KCCQ-OS score at enrollment in the CHAMP-HF trial

be expected to have a large and a very large health

was 63.6
23.7, corresponding to New York Heart Association functional class II. ASSOCIATION

BETWEEN

USE

status benefit, respectively, compared with the patient who had not been started on an ARNI therapy

AND

(Tables 3 and 4, Central Illustration). Moreover,

CHANGES IN HEALTH STATUS. Improvements in

OF

ARNI

fewer patients were likely to experience minimal to

KCCQ score from the last pre-match to the first post-

no improvement or deterioration in their health

match health status assessment were observed over

status

a median (IQR) of 57 (32, 104) days. Overall, ARNI

groups), translating to a NNT of 26 to prevent such

patients experienced an average 5.3
18.6-point

an outcome.

(absolute

difference

of

3.9%

between

improvement in the KCCQ-OS compared with 2.5


Consistent findings were observed among ARNI

17.4 points for their no-ARNI counterparts (differ-

patients who switched from ACE inhibitor/ARB ther-

ences in mean changes of 3.2 [95% CI: 1.5 to 4.9);

apy and among patients not previously taking an ACE

adjusted group-level differences with regression

inhibitor/ARB therapy (Online Figures 7 and 8). The

modeling of 2.9 points (95% CI: 1.14 to 4.6; p < 0.001)

proportions experiencing very large clinical im-

(Table 2). This was largely driven by patients’ im-

provements in the ACE inhibitor/ARB cohort were

provements in the KCCQ-PL (4.8
24.8 points vs. 2.0

19.1% versus 10.5%, respectively, suggesting that


22.2 points, respectively) and KCCQ-QoL (6.4
23.9

switching to an ARNI resulted in an NNT of 12 (95% CI:

points vs. 2.7
24.1 points, respectively) domains.

8 to 31) for a patient to experience a very large

Similar findings were observed in both the de novo

improvement in their health status. Among patients

ARNI initiates (mean difference of 2.9 points [95% CI:

not previously taking an ACE inhibitor/ARB, initiation

0.3 to 5.5; p ¼ 0.028]) and in those who switched

of the ARNI therapy was associated with a 22%

between ACE inhibitor/ARB and ARNI therapy (mean

(vs. 13.9%, respectively) chance of a very large

JACC: HEART FAILURE VOL.

-, NO. -, 2019

Khariton et al.

- 2019:-–-

improvement, corresponding to an NNT of 13 (95% CI: 8 to 50) (Table 4).

DISCUSSION Although a principal goal of HF management is to

Sacubitril/Valsartan and Health Status Outcomes

T A B L E 3 Patients Categorized by Pre-Post Changes in KCCQ Score in the

Combined Matched Cohort ARNI (n ¼ 508)

Pre-Post Change in KCCQ Score

Very large improvement ($20)

20.5 (104)

12.1 (123)

Large improvement (10 to <20)

12.2 (62)

14.8 (150)

alleviate patient symptoms and improve health sta-

Moderate improvement (5 to <10)

tus, few pharmacological therapies have been shown

No or small improvement or worse (<5)

to reliably reduce symptoms, improve functions, and

no-ARNI (n ¼ 1,016)

Overall summary

11.0 (56)

12.9 (131)

56.3 (286)

60.2 (612)

Physical limitation*

enhance quality of life. This study examined the early

Very large improvement ($20)

21.2 (101)

17.4 (160)

health status benefits of instituting an ARNI in a

Large improvement (10 to <20)

11.1 (53)

11.0 (101)

Moderate improvement (5 to <10)

11.6 (55)

10.4 (96)

56.1 (267)

61.2 (564)

multicenter observational registry of outpatients with chronic HFrEF. Independent of prior treatment with ACE inhibitor/ARB, it was found that patients pre-

No or small improvement or worse (<5) Symptom frequency Very large improvement ($20)

18.1 (92)

15.9 (162)

scribed ARNI experienced early and robust improve-

Large improvement (10 to <20)

13.0 (66)

12.1 (123)

ments in disease-specific health status as measured

Moderate improvement (5 to <10)

9.8 (50)

8.1 (82)

by the KCCQ. These findings represent the first real-

No or small improvement or worse (<5)

59.1 (300)

63.9 (649) 20.1 (192)

world evidence describing the potential health status benefits of ARNI in patients with HFrEF. These findings extend the clinical trial data from the PARADIGM-HF trial (11,19,20), which described significant reductions in cardiovascular mortality and

Social limitation† Very large improvement ($20)

23.9 (116)

Large improvement (10 to <20)

8.5 (41)

9.1 (87)

Moderate improvement (5 to <10)

9.9 (48)

12.1 (116)

57.7 (280)

58.6 (560)

No or small improvement or worse (<5) Quality of life‡

HFrEF hospitalization and greater preservation of

Very large improvement ($ 20)

25.6 (130)

22.3 (227)

health status over 8 months with ARNI. A notable

Large improvement (10 to <20)

20.7 (105)

16.9 (172)

limitation of PARADIGM-HF arose from the use of a

Moderate improvement (5 to <10)

run-in phase during which all patients underwent

No or small improvement or worse (<5)

dosages of enalapril followed by dose escalation of ARNI before baseline KCCQ assessment. As such, early improvements in health status associated

-

-

53.7 (273)

60.7 (617)

Values are % (n). *A total of 116 sets (348 patients) were omitted for missing pre- or post-match physical limitation scores for 1 or more members of the set. †A total of 76 sets (228 patients) were omitted for missing pre- or post-match social limitation scores for 1 or more members of the set. ‡Because of the granularity of the quality of life score, a “small improvement” was not possible.

with the use of ARNI could not be calculated. To address this gap, the CHAMP-HF registry offers a unique perspective by capturing patients’ true base-

to be treated for one to have a clinically important

line health status before ARNI initiation, thereby

change in their health status. The distribution of

providing a more accurate assessment of early health

changes in KCCQ scores for those started on ARNI

status changes over time after ARNI initiation.

was shifted, such that a substantial proportion of

Finding a clinically significant mean increase in

patients experienced very large clinical improve-

KCCQ-OS scores of 5.3 points in the patients treated

ments in their health status. It was found that an

with ARNI suggests that the early benefits of treat-

NNT of 12, meaning for every 12 patients treated

ment may have been missed in the PARADIGM-HF

with ARNI, compared with similar patients not

study but that, given the minimal further decrease

treated, 1 patient would experience a very large

in scores over time in the PARADIGM-HF, these ben-

improvement in their health status, regardless of

efits were likely sustained over time. Confirmation of

whether ARNI was used as a substitute for ACE

these sustained benefits from the CHAMP-HF study should be explored as more follow-up in this registry

T A B L E 4 Number Needed to Treat for Moderate and Large Thresholds of Health Status

accrues.

Change (Combined, ACE Inhibitor/ARB, and No-ACE Inhibitor/ARB Cohorts)

In comparing mean differences between groups, it can often be difficult to interpret the clinical signif-

Cohort

NNT for $Large KCCQ-OS Improvement (95% CI)

NNT for $Very Large KCCQ-OS Improvement (95% CI)

18 (10–111)

12 (9–24)

icance of changes. Accordingly, the distribution of

Combined

changes against well-established thresholds of clin-

ACE inhibitor/ARB*

-

12 (8–31)

ically important changes in the KCCQ were also

No-ACEI/ARB*

-

13 (8–50)

examined (15,17). By comparing the proportions of patients according to their magnitude of changes, a “responder” analysis could also be conducted and an estimate of the number of patients who would need

Values are NNT (confidence interval [CI]). *In the ACE inhibitor/ARB and No-ACE inhibitor/ARB cohorts, the number needed to treat (NNT) and 95% CIs for moderate or greater KCCQ overall summary (OS) improvement were not calculated due to lack of statistical significance. Other abbreviations as in Table 1.

7

Khariton et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Sacubitril/Valsartan and Health Status Outcomes

C E N T R A L IL LU ST R A T I O N Distribution of Changes in KCCQ Overall Summary Score (Combined Cohort)

40

31.3% 30.7%

30 Percent of Patients

8

20.5%

20 14.8% 12.9% 9.8%

10 6.3%

0

11.0%

7.6%

<–20

10.9%

11.0%

12.2%

12.1%

8.9%

–20 to <–10 –10 to <–5 –5 to 5 >5 to 10 >10 to 20 Change in KCCQ Summary Score (Post-Pre) ARNI

>20

No ARNI

Khariton, Y. et al. J Am Coll Cardiol HF. 2019;-(-):-–-. ARNI ¼ angiotensin-neprilysin inhibitor; KCCQ ¼ Kansas City Cardiomyopathy Questionnaire.

inhibitor/ARB treatment or as a de novo therapy.

findings may not be generalizable throughout the

This change of >20 points is comparable to the mean

United States or to other countries. In particular,

health status benefits after transcatheter aortic valve

although rates of medical therapy for heart failure

replacement or insertion of a left ventricular assist

were (at least) similar to those observed in prior

device (21,22). Understanding whether or not there

HFrEF registries and HFrEF clinical trials (23–25),

are particular patient profiles associated with such

rates of MRA, digoxin, and diuretic agents were lower

large health status benefits from ARNI is an impor-

than those in the PARADIGM-HF registry (25,26).

tant area for future investigation.

Finally, these results apply only to outpatients with

STUDY LIMITATIONS. First, the CHAMP-HF study

HFrEF, and the benefits of ARNI for other outcomes

was an observational registry and thereby susceptible

and in other heart failure populations require

to bias as a result of unmeasured confounding. In

further investigation.

particular, the role of placebo effect due to open-label ARNI use cannot be excluded. Second, patients’ lab-

CONCLUSIONS

oratory values were unable to be matched due to missing laboratory data. Third, although a primary

In an observational registry of outpatients with

aim of the CHAMP-HF study was to recruit partici-

HFrEF across the United States, this study found an

pating sites with diverse treatment backgrounds (e.g.,

association between ARNI initiation and early im-

physician specialty, patient population), the present

provements in patient-reported health status. These

JACC: HEART FAILURE VOL.

-, NO. -, 2019

Khariton et al.

- 2019:-–-

Sacubitril/Valsartan and Health Status Outcomes

improvements were largely driven by a substantially larger proportion of patients treated with ARNI who experienced a very large health status benefit shortly after ARNI initiation. These data supplement the benefits of ARNI therapy in reducing mortality and hospitalizations and underscore the need for future research to better identify patients who are most likely to benefit from ARNI.

PERSPECTIVES COMPETENCY IN MEDICAL KNOWLEDGE: Initiation of ARNI is associated with early, clinically meaningful health status (symptoms, functions, and quality of life) benefits in outpatients with heart failure and reduced ejection fraction. TRANSLATIONAL OUTLOOK: This work extends prior

ADDRESS FOR CORRESPONDENCE: Dr. Yevgeniy

Khariton, Saint Luke’s Mid America Heart Institute, 4401 Wornall Road, Kansas City, Missouri 64111.

insights into the clinical benefits of ARNI initiation and demonstrates its early health status benefits in a real-world HFrEF population.

E-mail: [email protected].

REFERENCES 1. Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the management of heart

results from the SHIFT study. Eur Heart J 2011;32: 2395–404.

failure: a report of the American College of Cardiology Foundation/American Heart Association task force on practice guidelines. J Am Coll Cardiol 2013;62:147–239.

11. McMurray JJ, Packer M, Desai AS, et al. Angiotensin-neprilysin inhibition versus enalapril in heart failure. N Engl J Med 2014;371:993–1004.

2. Rumsfeld JS, Alexander KP, Goff DC Jr., et al. Cardiovascular health: the importance of measuring patient-reported health status: a scientific statement from the American Heart Association. Circulation 2013;127:2233–49. 3. Pokharel Y, Khariton Y, Tang Y, et al. Association of serial Kansas City Cardiomyopathy Questionnaire assessments with death and hospitalization in patients with heart failure with preserved and reduced ejection fraction: a secondary analysis of 2 randomized clinical trials. JAMA Cardiol 2017;2: 1315–21. 4. Kosiborod M, Soto GE, Jones PG, et al. Identifying heart failure patients at high risk for nearterm cardiovascular events with serial health status assessments. Circulation 2007;115:1975–81. 5. Soto GE, Jones P, Weintraub WS, Krumholz HM, Spertus JA. Prognostic value of health status in patients with heart failure after acute myocardial infarction. Circulation 2004;110:546–51. 6. Spertus J, Peterson E, Conard MW, et al. Monitoring clinical changes in patients with heart failure: a comparison of methods. Am Heart J 2005;150:707–15.

12. Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the task force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail 2016;18:891–975. 13. Yancy CW, Jessup M, Bozkurt B, et al. 2016 ACC/AHA/HFSA focused update on new pharmacological therapy for heart failure: an update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Task Force on clinical practice guidelines and the Heart Failure Society of America. J Am Coll Cardiol 2016;68:1476–88. 14. DeVore AD, Thomas L, Albert NM, et al. Change the management of patients with heart failure: rationale and design of the CHAMP-HF registry. Am Heart J 2017;189:177–83. 15. Spertus JA, Jones PG. Development and validation of a short version of the Kansas City Cardiomyopathy Questionnaire. Circ Cardiovasc Qual Outcomes 2015;8:469–76.

7. Committee on Quality of Health Care in America Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001.

16. Flynn KE, Lin L, Moe GW, et al. Relationships between changes in patient-reported health status and functional capacity in outpatients with heart failure. Am Heart J 2012;163:88–94.

8. Kansas City Cardiomyopathy Questionnaire (KCCQ) j FDA Voice. Blogs.fda.gov. Available at: https://blogs.fda.gov/fdavoice/index.php/tag/kansascity-cardiomyopathy-questionnaire-kccq/. Accessed June 7, 2018.

17. Dreyer RP, Jones PG, Kutty S, Spertus JA. Quantifying clinical change: discrepancies between patients’ and providers’ perspectives. Qual Life Res 2016;25:2213–20.

Cures Act. Available at: https://www.nih.gov/ research-training/medical-research-initiatives/cures. Accessed June 14, 2018.

18. Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 2009;28:3083–107.

10. Ekman I, Chassany O, Komajda M, et al. Heart rate reduction with ivabradine and health related quality of life in patients with chronic heart failure:

19. Lewis EF, Claggett BL, McMurray JJV, et al. Health-related quality of life outcomes in PARADIGM-HF. Circ Heart Fail 2017;10:221.

9. National Institutes of Health. The 21st Century

20. Chandra A, Lewis EF, Claggett BL, et al. Effects of sacubitril/valsartan on physical and social activity limitations in patients with heart failure: a secondary analysis of the PARADIGM-HF trial. JAMA Cardiol 2018;3:498–505. 21. Osnabrugge RL, Arnold SV, Reynolds MR, et al. Health status after transcatheter aortic valve replacement in patients at extreme surgical risk: results from the CoreValve U.S. trial. J Am Coll Cardiol Intv 2015;8:315–23. 22. Rogers JG, Aaronson KD, Boyle AJ, et al. Continuous flow left ventricular assist device improves functional capacity and quality of life of advanced heart failure patients. J Am Coll Cardiol 2010;55:1826–34. 23. Fonarow GC, Yancy CW, Heywood JT, et al. Adherence to heart failure quality-of-care indicators in US hospitals: analysis of the ADHERE registry. Arch Intern Med 2005;165: 1469–77. 24. Patel DB, Shah RM, Bhatt DL, et al. Guidelineappropriate care and in-hospital outcomes in patients with heart failure in teaching and nonteaching hospitals: findings from get with the guidelines-heart failure. Circ Cardiovasc Qual Outcomes 2016;9:757–66. 25. McMurray JJ, Packer M, Desai AS, et al. Baseline characteristics and treatment of patients in prospective comparison of ARNI with ACE inhibitor to determine impact on global mortality and morbidity in heart failure trial (PARADIGM-HF). Eur J Heart Fail 2014;16:817–25. 26. Fonarow GC, Albert NM, Curtis AB, et al. Improving evidence-based care for heart failure in outpatient cardiology practices: primary results of the Registry to Improve the Use of EvidenceBased Heart Failure Therapies in the Outpatient Setting (IMPROVE HF). Circulation 2010;122: 585–96.

KEY WORDS health status, heart failure, sacubitril/valsartan A PP END IX For supplemental figures and tables, please see the online version of this paper.

9