- Email: [email protected]
Is Heart Rate Important for Patients With Heart Failure in Atrial Fibrillation?
Vol. -, No. -, 2014 ISSN 2213-1779/$36.00 http://dx.doi.org/10.1016/j.jchf.2014.01.005
JACC: Heart Failure 2014 by the American College of Cardiology Foundation Published by Elsevier Inc.
Is Heart Rate Important for Patients With Heart Failure in Atrial Fibrillation? Damien Cullington, MD, Kevin M. Goode, BENG, PHD, Jufen Zhang, PHD, John G. F. Cleland, MD, Andrew L. Clark, MA, MD Kingston upon Hull, United Kingdom Objectives
This study sought to investigate the relationship between resting ventricular rate and mortality in patients with chronic heart failure (CHF) and reduced left ventricular ejection fraction (LVEF) who were in sinus rhythm (SR) or atrial fibrillation (AF).
Background
Slower heart rates are associated with better survival in patients with CHF in SR, but it is not clear whether this is true for those in AF.
Methods
We assessed 2,039 outpatients with CHF and LVEF 50% undergoing baseline assessment, of whom 24% (n ¼ 488) were in AF; and 841 outpatients reassessed after attempted treatment optimization at 1 year, of whom 22% (n ¼ 184) were in AF. Cox proportional hazards models were used to assess the relationships between heart rate and survival in patients with CHF and AF or sinus rhythm. We analysed heart rate and rhythm data recorded at the baseline review and after 1-year follow-up. Proportional hazards assumptions were checked by Schoenfeld and Martingale residuals.
Results
The median survival for those in AF was 6.1 years (interquartile range [IQR]: 5.3 to 6.9 years) and 7.3 years (IQR: 6.5 to 8.1 years) for those in SR. In univariable analysis, patients with AF had a worse survival (hazard ratio [HR]: 1.26, 95% confidence interval [CI]: 1.08 to 1.47; p ¼ 0.003) but after covariate adjustment, survival rates were similar. After adjusting Cox regression models, there was no association between heart rate (per 10 beats/min increments) and survival in patients with AF before (HR: 0.94, 95% CI: 0.88 to 1.00, p ¼ 0.07) or after (HR: 1.00, 95% CI: 0.99 to 1.00, p ¼ 0.84) therapy optimization. For patients in SR, higher heart rates were associated with worse survival, both before (HR: 1.10, 95% CI: 1.05 to 1.15, p <0.0001) and after (HR: 1.13, 95% CI: 1.03 to 1.24, p ¼ 0.008) therapy optimization.
Conclusions
In patients with CHF and a reduced LVEF, slower resting ventricular rate is associated with better survival for patients in SR but not for those with AF. (J Am Coll Cardiol HF 2014;-:-–-) ª 2014 by the American College of Cardiology Foundation
Resting heart rate is a powerful prognostic marker in a broad range of subjects in sinus rhythm (SR), with or without cardiovascular disease (1–3). In patients with chronic heart failure (CHF) in SR due to left ventricular systolic dysfunction (LVSD), a lower resting heart rate is associated with better survival and fewer hospital readmissions due to heart failure (4,5). In contrast, a lower resting ventricular rate (<80 beats/min) is not associated with better survival in patients with permanent atrial fibrillation (AF), although studies have included few From the Department of Cardiology, Postgraduate Medical Institute, Hull York Medical School, University of Hull, Kingston upon Hull, United Kingdom. Dr. Cullington has received honoraria from Servier. Dr. Cleland has received research funding from Servier. Drs. Goode and Clark have received honoraria from Servier. Dr. Zhang has reported that he has no relationships relevant to the contents of this paper to disclose. Manuscript received December 17, 2012; revised manuscript received January 15, 2014, accepted January 17, 2014.
patients with CHF (6–8). Furthermore, there is little evidence that lower resting ventricular rate in patients with CHF and AF is important physiologically or improves quality of life (9–11). Accordingly, we investigated the relationship between ventricular rate and survival in patients with CHF due to LVSD who were in AF or SR. Methods Study population and data collection. We included outpatients with CHF who attended a community-based heart failure program in Kingston-Upon-Hull, United Kingdom, serving a population of 600,000 people, between September 1999 and October 2010. Patients were referred for diagnosis and/or management of CHF. Patients gave written consent for their data to be stored electronically and used for research purposes. A large proportion of surviving patients were reassessed at 1 year after attempts had
2
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
been made to optimize therapy in accordance with UK guidelines for the management of ACE-I = angiotensinchronic heart failure (12). converting enzyme inhibitor All patients underwent a AF = atrial fibrillation standardized clinical examinaARB = angiotensin receptor tion, 12-lead electrocardiography, blocker echocardiography, spirometry, CHF = chronic heart failure and routine blood tests. Cardiac DBP = diastolic blood rhythm and resting heart rate pressure were determined from a 12-lead HR = heart rate electrocardiogram, which was IHD = ischemic heart obtained after at least 5 min rest in disease the supine position, using a MAC IQR = interquartile range 5000 machine (GE Healthcare, LBBB = left bundle branch Milwaukee, Wisconsin). We used block the arbitrary definitions of the LVSD = left ventricular RACE II (Rate Control Efficacy systolic dysfunction in Permanent Atrial Fibrillation: NYHA = New York Heart a Comparison between Lenient Association versus Strict Rate Control II) SBP = systolic blood study to categorize patients in AF pressure into those with “lenient” (80 SR = sinus rhythm and <110 per min) or “strict” (<80 per min) ventricular rate control (7). Patients who were permanently paced were excluded from the analysis. All echocardiograms were performed by experienced echocardiographers in accordance with recommendations of the British Society of Echocardiography. For echocardiogram acquisition, a Vivid 5 or Vivid 7 system (GE Healthcare) using a 3.4-MHz probe was used. When possible, the Simpson’s biplane method was used to measure the left ventricular ejection fraction (LVEF), and, in all cases, a visual approximation was made. Only patients with at least mild left ventricular systolic dysfunction (LVSD) were included. The primary endpoint of the study was all-cause mortality. Mortality data were captured electronically by National Health Service information systems and reported to the hospital and were complete to the censor date of August 31, 2011. Ethics. The investigation conformed to the principles outlined in the Declaration of Helsinki. It was approved by the Hull and East Yorkshire Research Ethics Committee (Heart Care Study ELSY 2642). All subjects gave written informed consent. Statistical methods. Data were tested for normality using the Kolmogorov-Smirnov test. Normally distributed data are presented as mean SD; other data are presented as medians and interquartile ranges (IQR). Categorical data are given as percents. Differences in categorical variables between groups were compared with the chi-square test. Differences between groups with normally distributed data were analyzed using Student’s t test, and non-normally distributed data were analyzed using the Mann-Whitney U test. Abbreviations and Acronyms
Kaplan-Meier curves used for analysis of survival and significance tests between groups were tested using the logrank test. Univariate and multivariate Cox regression models were constructed to investigate the relationship between potential covariates and outcome. To avoid overfitting, we included only variables significantly associated with outcome in the univariate analysis (p < 0.05) in the forward conditional stepwise Cox analysis. Proportional hazards assumptions were checked by Schoenfeld residuals, and Martingale residuals plots were used to evaluate linearity (13,14). Results are presented as hazard ratios (HRs) with 95% confidence intervals (CIs). Analyses were performed using SPSS version 19 (IBM Corp., Armonk, New York) and Stata 13 software (StataCorp., College Station, Texas). Results Baseline patient characteristics. A total of 2,039 patients was included in the baseline analysis, of whom 24% (n ¼ 488) had AF (Table 1). Compared to patients in SR, patients in AF were older and less likely to be diabetic or to have ischemic heart disease. Although there were no differences in severity of LVSD, patients with AF were more symptomatic than patients in SR. Patients with AF were less likely to be prescribed a beta-blocker, aldosterone antagonist, statin, or aspirin but more likely to be prescribed warfarin, digoxin, or a loop diuretic agent. The median resting heart rate was higher for patients with AF than for patients with SR (81 per min [IQR: 69 to 98] versus 70 (IQR: 60 to 83) per min, respectively, p ¼ 0.0001). Relationship between heart rhythm and survival. During the follow-up period, 229 patients (47%) with AF died compared to 639 patients (41%) with SR. The 1-year mortality rate for patients with AF was 15% (n ¼ 74) versus 11% for patients with SR (n ¼ 170). Overall median follow-up was 3.6 years (IQR: 1.7 to 6.9 years). The overall median survival for patients with AF was 6.1 years (IQR: 5.3 to 6.9 years) compared to 7.3 years (IQR: 6.5 to 8.1 years) for patients with SR. In a multivariable model using baseline data that was corrected for age, sex, weight, QRS duration, heart rate, systolic blood pressure, New York Heart Association (NYHA) class III/IV versus I/II, ischemic heart disease (IHD) status (yes/no), diabetic status (yes/no), angiotensinconverting enzyme/angiotensin receptor blocker (ACE/ ARB) use (yes/no), LVSD severity (moderate vs.
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Table 1
3
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
Characteristics of Patients in AF and SR at Baseline and at 1-Year Follow-Up Baseline Review % of Patients in AF (n ¼ 488)
Age, yrs Male Weight, kg
% of Total Patient Cohort
“Strict” HR (n ¼ 219)
“Lenient” HR (n ¼ 200)
HR 110 beats/min (n ¼ 69)
76 (70–81)
74 (68–80)
72 (63–78)
75
76
80 (67–92)
83 (69–97)
Patients in AF (n ¼ 488)
Patients In SR (n ¼ 1,551)
p Value
0.01
74 (68–80)
70 (63–77)
0.0001
64
0.11
74
72
0.45
84 (69–97)
0.25
81.6 (68–95)
78.9 (67–91)
0.01
0.04
15
19
0.03
0.0001
53
69
Diabetic patient
15
19
6
IHD
61
52
32
NYHA functional class
p Value
I
11
7
9
9
17
II
53
45
45
49
51
III
35
43
41
40
29
IV
1
3
6
2
2
SBP, mm Hg
130 (112–147)
127 (116–146)
122 (110–136)
128 (112–145)
130 (115–147)
DBP, mm Hg
74 (65–87)
80 (70–90)
84 (76–96)
79 (68–90)
75 (66–83)
Mild
23
17
16
19
17
Mild-to-moderate
14
18
13
15
12
Moderate
25
18
7
20
23
Moderate-to-severe
21
22
26
22
23
Severe
17
26
38
24
25
67 (59–73)
90 (85–98)
121 (113–132)
0.0001
81 (69–98)
70 (60–83)
112 (98–138)
104 (92–120)
102 (90–116)
0.0001
106 (96–128)
110 (96–136)
HR, beats/min (IQR)
0.0001 0.003
LVSD
QRS duration, ms (IQR)
0.12
0.0001 0.10 0.02
None
39
51
67
48
Low dose
29
27
22
32
34
Moderate dose
24
18
12
15
19
High dose
0.09 0.0001 0.18
0.002
Beta-blocker*
0.0001 0.0001
0.09
40
7
5
0
5
7
ACE-I/ARB
82
68
77
0.003
76
79
0.07
Aldosterone antagonist
25
20
7
0.005
20
25
0.03
Loop diuretic agent
83
77
81
0.29
80
71
0.0001
Digoxin
51
51
48
0.88
51
8
0.0001
Amiodarone
10
8
7
0.69
8
8
0.70
1
5
7
0.03
3
3
0.72
Statin
41
35
23
0.03
36
57
0.0001
Warfarin
38
49
13
0.06
58
12
0.0001
Aspirin
50
38
13
0.06
29
59
0.0001
5
5
3
0.75
13
5
0.0001
Verapamil/diltiazem
Clopidogrel
Continued on the next page
with a 7% decrease in mortality per year (HR: 0.93; 95% CI: 0.88 to 0.99, p ¼ 0.02) (Table 2). After multivariable adjustment, heart rate as a continuous variable was not significantly associated with survival (HR: 0.94; 95% CI: 0.88 to 1.00, p ¼ 0.07), nor was it when split by quartiles (Fig. 1). For patients in SR, in the multivariable model, an increase in resting heart rate by 10 beats/min was associated with a 10% increase in mortality per year (HR: 1.10; CI: 95% 1.04 to 1.15, p <0.0001]. The group was divided into quartiles according to heart rate: <60/min (referent); 60 to 70/min; 71 to 83/min, and >83/min. Compared to the referent group, only patients with a resting heart rate >83/min had
significantly worse survival (HR: 1.69; CI: 1.30 to 2.21, p ¼ 0.0001). Patient characteristics at 1 year. A total of 841 patients was seen after the first year of active treatment in the clinic, of whom 22% (n ¼ 184) were in AF (Table 1). Patients who had AF were older and weighed more but otherwise had similar characteristics to patients in SR. Patients with AF had higher resting heart rate than patients in SR (72 beats/min [IQR: 63 to 82]) vs. 63 beats/min [IQR: 57 to 71], respectively; p ¼ 0.0001). Prescription of betablockers, ACE inhibitors/ARBs, or aldosterone antagonists was similar in the 2 groups. Patients in AF were more likely to be prescribed loop diuretic agents, digoxin, and
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
4
Table 1
Continued 1- Year Review % of Patients in AF (n ¼ 184)
% of Total Patient Cohort
“Strict” HR (n ¼ 129)
“Lenient” HR (n ¼ 48)
HR 110 beats/min (n ¼ 7)
73 (67–79)
75 (64–80)
77 (66–79)
82
79
85 (74–97)
Patients in AF (n ¼ 184)
Patients in SR (n ¼ 657)
p Value
0.85
73 (66–79)
69 (62–75)
0.0001
57
0.26
80
75
0.14
83 (69–96)
73 (69–100)
0.43
84.6 (73–97)
81.1 (70–92)
0.03
19
28
17
0.45
20
18
0.38
66
60
57
0.74
69
64
p Value
0.01
0.25 0.06
12
13
43
13
22
71
52
14
64
60
16
33
43
22
17
2
2
0
2
1
124 (110–141)
127 (112–151)
116 (103–134)
0.37
125 (110–141)
126 (115–143)
0.16
72 (65–83)
80 (68–92)
65 (62–86)
0.06
73 (65–84)
72 (64–80)
0.08
24
29
14
25
22
15
6
43
14
14
19
23
14
20
23
27
17
14
24
24
15
25
14
17
17
65 (59–73)
85 (83–91)
120 (114–126)
0.0001
72 (63–82)
63 (57–71)
114 (102–144)
110 (94–134)
90 (84–96)
0.0001
114 (98–142)
110 (96–138)
0.17
0.91
0.005
0.0001 0.44 0.75
11
27
29
16
14
41
19
57
36
34
33
25
14
30
34
16
29
0
18
18
94
94
100
0.79
94
92
0.31
35
23
29
0.31
32
28
0.35
88
85
100
0.58
88
74
0.0001
55
52
43
0.79
54
10
0.0001
12
10
0
0.58
11
10
0.47
1
4
0
0.27
2
2
0.86
43
31
57
0.25
40
66
0.0001
20
23
29
0.82
78
18
0.0001
83
67
71
0.06
21
55
0.0001
4
0
14
0.11
3
12
0.0001
Values are median (interquartile range) or n. *Doses of beta-blocker: None, 0%; Low dose, 1% to 33%; Moderate dose, 34% to 66%; High dose, 67% to 100%. Values of p 0.05 are in bold. ACE-I/ARB ¼ angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; AF ¼ atrial fibrillation; DBP ¼ diastolic blood pressure; HR ¼ heart rate; IHD ¼ ischemic heart disease; IQR ¼ interquartile range; LBBB ¼ left bundle branch block; LVSD ¼ left ventricular systolic dysfunction; NYHA ¼ New York heart association; SBP ¼ systolic blood pressure; SR ¼ sinus rhythm.
warfarin but less likely to be prescribed statins, aspirin, or clopidogrel. Relationship between heart rate at 1 year and survival. In univariable analysis restricted to the patients in AF, ventricular rate was not a significant predictor of survival (HR: per 10 beats/min increase in heart rate, 1.01; 95% CI: 0.90 to 1.13, p ¼ 0.84) (Fig. 2; Table 3). In contrast, for patients in SR, heart rate was significantly associated with survival. In the multivariable model, each increase in resting heart rate by 10 beats/min was associated with an 13% increase in annual mortality (HR: 1.13, 95%
CI: 1.03 to 1.24, p ¼ 0.008). The group was subdivided into quartiles according to heart rate: <57 beats/min; 57 to 62 beats/min; 63 to 70 beats/min; and >70 beats/min. Only patients with a resting heart rate >70 beats/min had worse survival than referent group (<57 beats/min; HR: 1.47; 95% CI: 1.03 to 2.13, p ¼ 0.04). Discussion This analysis suggests that after adjusting for age and other differences, patients with CHF and LVSD who are
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Table 2
Univariable and Multivariable Analyses of Patients in AF and SR at Baseline* Patients in AF HR Represents
Age
Per decade increase
Patients in SR
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
1.65 (1.41–1.93)
<0.0001
1.53 (1.29–1.82)
<0.0001
1.76 (1.61–1.92)
<0.0001
1.73 (1.58–1.90)
<0.0001
Sex
Men vs. women
1.09 (0.81–1.45)
0.59
–
–
1.26 (1.06–1.49)
0.008
–
–
Weight
Per 5-kg increase
0.95 (0.92–0.98)
0.004
–
–
0.92 (0.90–0.94)
<0.0001
–
–
1.05 (1.005–1.10)
QRS duration
Per 10-ms increase
1.07 (1.02–1.12)
0.004
1.03 (1.00–1.05)
0.04
–
–
LVI severity
Moderate vs.
1.30 (0.97–1.73)
0.08
–
–
1.30 (1.08–1.56)
0.005
–
–
Heart rate
Per 10-beats/min increase
0.93 (0.88–0.99)
0.02
–
–
1.15 (1.10–1.20)
<0.0001
1.10 (1.05–1.15)
<0.0001
Systolic BP
Per 5-U increase
0.97 (0.94–1.00)
0.03
–
–
0.97 (0.96–0.99)
0.001
0.97 (0.96–0.99)
0.002
NYHA class
III/IV vs. I/II
1.82 (1.40–2.38)
<0.0001
1.51 (1.14–1.99)
0.004
1.80 (1.53–2.12)
<0.0001
1.32 (1.11–1.57)
0.002
IHD
Yes vs. no
1.63 (1.25–2.13)
<0.0001
1.53 (1.14–2.04)
0.005
1.09 (0.92–1.30)
0.32
Diabetic patient
Yes vs. no
1.15 (0.80–1.64)
–
–
1.31 (1.09–1.59)
0.005
0.45
0.03
–
–
1.36 (1.11–1.66)
0.003
Yes vs. no
0.86 (0.65–1.15)
0.32
–
–
0.78 (0.65–0.93)
0.006
0.80 (0.66–0.97)
0.021
Yes vs. no
1.34 (0.94–1.89)
0.10
–
–
2.21 (1.81–2.70)
<0.0001
1.69 (1.35–2.11)
<0.0001 –
0.22
–
Aldosterone Antagonist
Yes vs. no
1.22 (0.89–1.69)
Beta-blocker
Yes vs. no
0.67 (0.516–0.874)
Digoxin
Yes vs. no
0.92 (0.71–1.19)
0.50
–
Amiodarone
Yes vs. no
1.27 (0.83–1.93)
0.28
–
0.003
0.71 (0.53–0.94)
–
1.29 (1.08–1.53)
0.005
–
0.71 (0.60–0.82)
<0.0001
–
–
–
1.48 (1.14–1.93)
0.003
1.50 (1.14–1.98)
0.004
–
1.31 (1.01–1.70)
0.046
1.37 (1.04–1.80)
0.026
0.02
*Hazard ratios for primary endpoint mortality. yVariables retained in the multivariate Cox model are shown. Values in bold indicate hazard ratios with a value of p 0.05. ACE/ARB ¼ angiotensin-converting enzyme/angiotensin blocker receptor; AF ¼ atrial fibrillation; BP ¼ blood pressure; CI ¼ confidence interval; HR ¼ hazard ratio; IHD ¼ ischemic heart disease; LVSD ¼ left ventricular systolic dysfunction; MVM ¼ multivariate model; NYHA ¼ New York Heart Association; SR ¼ sinus rhythm; UVM ¼ univariate model.
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
ACE/ARB Loop diuretic agent use
5
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
6
Figure 1
Multivariable Adjusted Survival Curves for Patients in Atrial Fibrillation at Baseline Divided by Heart Rate Quartiles
CI ¼ confidence interval; HR ¼ hazard ratio.
in AF or SR have a similar prognosis, despite patients in AF having a substantially higher ventricular rate. After 1 year of follow-up and optimization of medical therapy, resting ventricular rate was <80 beats/min in 70% of patients with AF and the median ventricular rate was only 72 beats/min (IQR: 63 to 83 beats/min), although this remained somewhat higher than the median rate for patients in SR (63 beats/min; IQR: 57 to 71 beats/min). Thus, adequate rate control might have contributed to the similar survival of patients in AF and SR. In contrast to patients in SR, higher ventricular rates were not associated with increased mortality in patients with CHF, LVSD, and AF. Similar conclusions have been found from post-hoc analysis of the CHARM (Candesartan in Heart Failure: Assessment of Reduction in Mortality and morbidity) studies (15). There is a strong association between ventricular rate and mortality for patients with CHF in SR (4). If this is solely a rate-mediated effect, then a similar relationship would be expected for patients in AF. However, a growing body of evidence suggests that ventricular rate may be less important among patients with AF (8,9,11). Pooled data from the AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) and
Figure 2
Multivariable Adjusted Survival Curves for Patients in Atrial Fibrillation after 1-Year Follow-Up Divided by Heart Rate Quartiles
Abbreviations as in Figure 1.
RACE studies suggest that chronic resting heart rates >100 beats/min might be detrimental (16). Current recommendations suggest that resting ventricular rate should be lowered to <100 beats/min for the asymptomatic patient and, more aggressively (i.e., <80 beats/min at rest and <110 beats/min with moderate exercise) if symptoms persist or if LVEF deteriorates (17,18). Some patients with AF and CHF have better symptom control and higher LVEF when a lower ventricular rate is achieved by using a combination of carvedilol and digoxin (10). The RACE II study (n ¼ 614), is the only prospectively designed randomized controlled trial to investigate whether “strict” or “lenient” ventricular rate control should be preferred in patients with AF (7). Neither strategy was associated with any significant difference in mortality, hospitalizations, or symptoms during 3-year follow-up. Only 10% of the patients enrolled had a history of heart failure, and the mean LVEF was 51%, thus, extrapolation of the data to steer the management of patients with CHF would seem inappropriate. Clearly, there must be a ventricular rate below which harm occurs. The notion that sudden cardiac death in patients with CHF is largely due to tachyarrhythmia may be exaggerated. The CARISMA (Cardiac Arrhythmias
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Table 3
Univariable and Multivariable Analyses of Patients in AF and SR at 1 Year* Patients in AF HR Represents
Patients in SR
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
1.55 (1.21–1.99)
0.001
2.16 (1.88–2.50)
<0.0001
2.01 (1.75–2.33)
<0.0001
Age
Per decade increase
1.51 (1.19–1.91)
0.001
Sex
Men vs. women
1.004 (0.62–1.62)
0.99
–
–
1.25 (0.96–1.63)
–
–
Weight
Per 5-kg increase
0.96 (0.91–1.01)
0.15
–
–
0.91 (0.88–0.95)
<0.0001
–
– 0.035
0.09
Per 10-ms increase
1.05 (0.99–1.13)
0.11
–
–
1.05 (1.03–1.08)
<0.0001
1.03 (1.00–1.06)
Moderate vs.
1.65 (1.09–2.49)
0.02
1.61 (1.07–2.45)
0.024
1.56 (1.21–2.02)
0.001
1.36 (1.04–0.78)
0.025
Heart rate
Per 10-beats/min increase
1.01 (0.90–1.13)
0.84
–
–
1.13 (1.03–1.23)
0.008
1.13 (1.03–1.24)
0.008
Systolic BP
Per 5-U increase
1.00 (0.96–1.04)
0.95
–
–
1.00 (0.97–1.02)
NYHA class
III/IV vs. I/II
1.53 (0.99–2.35)
0.05
–
–
2.26 (1.73–2.95)
IHD
Yes vs. no
1.10 (0.73–1.65)
0.66
–
–
0.96 (0.75–1.24)
Diabetic patient
Yes vs. no
1.41 (0.89–2.23)
0.15
–
–
1.30 (0.97–1.74)
ACE/ARB use
Yes vs. no
0,96 (0.42–2.20)
0.93
–
–
0.67 (0.46–0.98)
Loop diuretic use
Yes vs. no
2.47 (1.15–5.33)
0.02
2.48 (1.15–5.36)
0.021
2.10 (1.53–2.87)
Aldosterone antagonist use
Yes vs. no
0.90 (0.58–1.38)
0.61
–
–
1.20 (0.93–1.55)
Beta-blocker use
Yes vs. no
0.83 (0.49–1.40)
0.48
–
–
0.53 (0.40–0.71)
–
–
1.68 (1.26–2.25)
<0.0001
0.76
–
–
0.08
–
–
0.04
–
–
1.42 (1.02–1.99)
0.039
0.77 <0.0001
<0.0001
–
–
<0.0001
–
–
0.17
Digoxin use
Yes vs. no
0.91 (0.62–1.35)
0.65
–
–
1.74 (1.17–2.57)
0.006
–
–
Amiodarone use
Yes vs. no
1.24 (0.69–2.22)
0.48
–
–
1.13 (0.77–1.66)
0.54
–
–
*Hazard ratios for the primary endpoint mortality. yVariables retained in the multivariate Cox model are shown. Values in bold indicate hazard ratios with a value of p 0.05. Abbreviations as in Table 2.
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
QRS duration LVSD severity
7
8
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
and Risk Stratification After Acute Myocardial Infarction) trial analyzed data over 2 years from implantable loop recorders in 297 patients with IHD and a reduced LVEF (19). After multivariate adjustment, high degree atrioventricular block was the strongest predictor of cardiac death (HR: 6.75, 95% CI: 2.55 to 17.84, p < 0.001), followed by sinus bradycardia (HR: 4.15, 95% CI: 1.37 to 12.62, p ¼ 0.012). The number of patients enrolled with AF was small (n ¼ 8 permanent and n ¼ 17 paroxysmal), but half of the patients with permanent, and 12% with paroxysmal AF went on to develop high-degree atrioventricular block compared to only 8% of patients in SR (20). Similarly, patients in strict heart rate control groups in the AFFIRM and RACE trials (most of whom did not have LVSD) had a higher incidence of pacemaker implantation (16). In a substudy of PRIME II (Second Prospective Randomised Study of Ibopamine on Mortality and Efficacy), patients with AF and advanced heart failure with a baseline resting heart rate <80 beats/min had worse survival (HR: 2.9, 95% CI: 1.4 to 5.8, p ¼ 0.002) than patients with heart rates >80 beats/min (21). Atrial fibrillation causes loss of “atrial kick,” reduced LV diastolic filling, and a fall in stroke volume. An increase in ventricular rate may compensate for these changes to maintain adequate cardiac output; therefore, aggressive rate lowering in patients with CHF and AF may be hemodynamically detrimental (22). Tight daytime control of ventricular rate may, for selected patients, be associated with an exacerbation of nocturnal pauses which may increase the likelihood of pause-dependent ventricular tachycardia (10,23). The prevalence of AF in patients with CHF is estimated to be 30% to 40%, which increases with worsening LVSD (24). The baseline prevalence of AF in randomized trials of betablockers to treat CHF ranges from 13% to 35%. Bisoprolol, metoprolol, and nebivolol have not been shown to improve outcomes in patients with AF and CHF, although carvedilol may (25,26). If beta-blockers are less effective or ineffective in patients with heart failure and AF, this may reflect the lack of importance of ventricular rate control or that exacerbation of nocturnal pauses detracts from the benefits of rate control. Settling this issue is important for future clinical practice. Study limitations. This was a post-hoc analysis of prospectively collected data with all the limitations that imposes. Resting heart rate is a simple but poor indicator of heart rate control during activity, and we appreciate that a better assessment of risk might be obtained from 24-h monitoring. The analysis does not control for any medical treatment changes or pacemaker implantation after 1 year of follow-up. We sampled patients at 2 time points and did not control for patients who switched rhythms, although such patients were few. Atrial fibrillation may be an independent predictor of an adverse outcome in patients with CHF and preserved ejection fraction, but we did not explore this because follow-up data collection was restricted to patients with LVSD. We could not define whether beta-blockers were associated with a
lower mortality in our observational study, but the analysis may have lacked power and could not adjust for unmeasured confounders. Conclusions For patients with chronic heart failure in SR, lower heart rates are associated with lower mortality, but this is not the case for similar patients in AF. Whether it is advantageous to lower heart rate in CHF patients with AF to a range similar to that recommended for patients in sinus rhythm (<70 beats/min) is uncertain and requires further investigation. The benefits of daytime ventricular rate reduction may be balanced by an increased risk, possibly due to exacerbation of nocturnal pauses.
Reprint requests and correspondence: Dr. Damien Cullington, Yorkshire Heart Centre, Department of Adult Congenital Heart Disease, Jubilee Building, Leeds General Infirmary, Leeds, West Yorkshire LS1 3EX, United Kingdom. E-mail: [email protected].
REFERENCES
1. Diaz A, Bourassa MG, Guertin MC, Tardif JC. Long-term prognostic value of resting heart rate in patients with suspected or proven coronary artery disease. Eur Heart J 2005;26:967–74. 2. Palatini P, Thijs L, Staessen JA, et al., for the Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Predictive value of clinic and ambulatory heart rate for mortlality in elderly subjects with systolic hypertension. Arch Intern Med 2002;162:2313–21. 3. Dyer AR, Persky V, Stamler J, et al. Heart rate as a prognostic factor for coronary heart disease and mortality : findings in three Chicago epidemiological studies. Am J Epidemiol 1980;112:737–49. 4. Böhm M, Swedberg K, Komajda M, Borer JS. Heart rate as a risk factor in chronic heart failure (SHIFT): the association between heart rate and outcomes in a randomised placebo-controlled trial. Lancet 2010;376:886–94. 5. Böhm M, Borer J, Ford I, et al. Heart rate at baseline influences the effect of ivabradine on cardiovascular outcomes in chronic heart failure: analysis from the SHIFT study. Clin Res Cardiol 2013;102:11–22. 6. Heist EK, Mansour M, Ruskin JN. Rate control in atrial fibrillation. Circulation 2011;124:2746–55. 7. Van Gelder IC, Groenveld HF, Crijns HJ, et al., for the RACE II Investigators. Lenient versus strict rate control in patients with atrial fibrillation. N Engl J Med 2010;362:1363–73. 8. Cooper HA, Bloomfield DA, Bush DE, et al., for the AFFIRM Investigators. Relation between achieved heart rate and outcomes in patients with atrial fibrillation (from the Atrial Fibrillation Follow-up Investigation of Rhythm Management [AFFIRM] Study). Am J Cardiol 2004;93:1247–53. 9. Groenveld HF, Crijns HJ, Van den Berg MP, et al., for the RACE II Investigators. The effect of rate control on quality of life in patients with permanent atrial fibrillation. J Am Coll Cardiol 2011;58: 1798–803. 10. Khand AU, Rankin AC, Martin W, Taylor J, Gemmell I, Cleland JGF. Carvedilol alone or in combination with digoxin for the management of atrial fibrillation in patients with heart failure. J Am Coll Cardiol 2003;42:1944–51. 11. Smit MD, Crijns HJ, Tijssen JG, et al., for the RACE II Investigators. Effect of lenient versus strict rate control on cardiac remodelling in patients with atrial fibrillation. J Am Coll Cardiol 2011;58:942–9. 12. National Institute for Health and Care Institute. Chronic Heart Failure. National Clinical Guideline for Diagnosis and Management in
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
13. 14. 15.
16.
17.
18. 19.
Primary and Secondary Care. Available at: www.nice.org.uk/CG108. Accessed June 2012. Grambsch PM, Therneau TM. Proportional hazards and diagnostics based on weighted residuals. Biometrika 1994;81:515–26. Therneau TM, Grambsch PM, Fleming TR. Martingale based residuals for survival models. Biometrika 1990;77:147–60. Castagno D, Skali H, Takeuchi M, et al., for the CHARM Investigators. Association of heart rate and outcomes in a broad spectrum of patients with chronic heart failure: results from the CHARM (Candesartan in Heart Failure: Assessment of Reduction in Mortality and morbidity) program. J Am Coll Cardiol 2012;59: 1785–95. Van Gelder IC, Wyse DG, Chandler ML, et al., for the RACE and AFFIRM Investigators. Does intensity of rate-conrtol influence outcome in atrial fibrillation? An analysis of pooled data from the RACE and AFFIRM studies. Europace 2006;8:935–42. Gillis AM, Verma A, Talajic M, Nattel S, Dorian Pfor the CCS Atrial Fibrillation Guidelines Committee. Canadian Cardiovascular Society atrial fibrillation guidelines 2010: rate and rhythm management. Can J Cardiol 2011;27:47–59. Wasmer K, Eckardt L. Management of atrial fibrillation around the world: a comparison of current ACCF/AHA/HRS/CCS, and ESC guidelines. Europace 2011;13:1368–74. Bloch Thomsen PE, Jons C, Raatikainen MJ, et al., for the Cardiac Arrhythmias and Risk Stratification After Acute Myocardial Infarction (CARISMA) Study Group. Long-term recording of cardiac arrhythmias with an implantable cardiac monitor in patients with reduced ejection fraction after acute myocardial infarction. The Cardiac Arrhythmias and Risk Stratification After Acute Myocardial Infaction (CARISMA) Study. Circulation 2010;121:1258–64.
9
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
20. Gang UJ, Jøns C, Jørgensen RM, et al., for the CARISMA investigators. Risk markers of late high-degree atrioventricular block in patients with left ventricular dysfuction after an acute myocardial infarction: a CARISMA substudy. Europace 2011;13:1471–7. 21. Rienstra M, Van Gelder IC, Van den Berg MP, Boomsma F, Hillege HL, Van Veldhuisen DJ. A comparison of low versus high heart rate in patients with atrial fibrillation and advanced chronic heart failure: effects on clinical profile, neurohormonaes and survival. Int J Cardiol 2006;109:95–100. 22. Rawles JM. What is meant by a “controlled” ventricular rate in atrial fibrillation? Br Heart J 1990;63:157–61. 23. Sweeney MO, Ruetz LL, Belk P, Mullen TJ, Johnson JW, Sheldon T. Bradycardia pacing-induced short-long-short sequences at the onset of ventricular tachyarrhythmias: a possible mechanism of proarrhythmia? J Am Coll Cardiol 2007;50:614–22. 24. Maggioni AP, Dahlström U, Filippatos G, et al., for the Heart Failure Association of European Society of Cardiology. EURObservational research program: the Heart Failure Pilot Survey (ESC-HF Pilot). Eur J Heart Fail 2010;12:1076–84. 25. Joglar JA, Acusta AP, Shusterman NH, et al. Effect of carvedilol on survival and haemodynamics in patients with atrial fibrillation and left ventricular dysfunction: retrospective analysis of the US carvedilol heart failure trials program. Am Heart J 2001;142:498–501. 26. Rienstra M, Damman K, Mulder BA, Van Gelder IC, McMurray JJV, Van Veldhuisen DJ. Beta-blockers and outcome in heart failure and atrial fibrillation: a meta-analysis. J Am Coll Cardiol HF 2013;1:21–8. Key Words: atrial fibrillation sinus rhythm.
-
heart failure
-
heart rate
-
prognosis
-
JACC: Heart Failure 2014 by the American College of Cardiology Foundation Published by Elsevier Inc.
Is Heart Rate Important for Patients With Heart Failure in Atrial Fibrillation? Damien Cullington, MD, Kevin M. Goode, BENG, PHD, Jufen Zhang, PHD, John G. F. Cleland, MD, Andrew L. Clark, MA, MD Kingston upon Hull, United Kingdom Objectives
This study sought to investigate the relationship between resting ventricular rate and mortality in patients with chronic heart failure (CHF) and reduced left ventricular ejection fraction (LVEF) who were in sinus rhythm (SR) or atrial fibrillation (AF).
Background
Slower heart rates are associated with better survival in patients with CHF in SR, but it is not clear whether this is true for those in AF.
Methods
We assessed 2,039 outpatients with CHF and LVEF 50% undergoing baseline assessment, of whom 24% (n ¼ 488) were in AF; and 841 outpatients reassessed after attempted treatment optimization at 1 year, of whom 22% (n ¼ 184) were in AF. Cox proportional hazards models were used to assess the relationships between heart rate and survival in patients with CHF and AF or sinus rhythm. We analysed heart rate and rhythm data recorded at the baseline review and after 1-year follow-up. Proportional hazards assumptions were checked by Schoenfeld and Martingale residuals.
Results
The median survival for those in AF was 6.1 years (interquartile range [IQR]: 5.3 to 6.9 years) and 7.3 years (IQR: 6.5 to 8.1 years) for those in SR. In univariable analysis, patients with AF had a worse survival (hazard ratio [HR]: 1.26, 95% confidence interval [CI]: 1.08 to 1.47; p ¼ 0.003) but after covariate adjustment, survival rates were similar. After adjusting Cox regression models, there was no association between heart rate (per 10 beats/min increments) and survival in patients with AF before (HR: 0.94, 95% CI: 0.88 to 1.00, p ¼ 0.07) or after (HR: 1.00, 95% CI: 0.99 to 1.00, p ¼ 0.84) therapy optimization. For patients in SR, higher heart rates were associated with worse survival, both before (HR: 1.10, 95% CI: 1.05 to 1.15, p <0.0001) and after (HR: 1.13, 95% CI: 1.03 to 1.24, p ¼ 0.008) therapy optimization.
Conclusions
In patients with CHF and a reduced LVEF, slower resting ventricular rate is associated with better survival for patients in SR but not for those with AF. (J Am Coll Cardiol HF 2014;-:-–-) ª 2014 by the American College of Cardiology Foundation
Resting heart rate is a powerful prognostic marker in a broad range of subjects in sinus rhythm (SR), with or without cardiovascular disease (1–3). In patients with chronic heart failure (CHF) in SR due to left ventricular systolic dysfunction (LVSD), a lower resting heart rate is associated with better survival and fewer hospital readmissions due to heart failure (4,5). In contrast, a lower resting ventricular rate (<80 beats/min) is not associated with better survival in patients with permanent atrial fibrillation (AF), although studies have included few From the Department of Cardiology, Postgraduate Medical Institute, Hull York Medical School, University of Hull, Kingston upon Hull, United Kingdom. Dr. Cullington has received honoraria from Servier. Dr. Cleland has received research funding from Servier. Drs. Goode and Clark have received honoraria from Servier. Dr. Zhang has reported that he has no relationships relevant to the contents of this paper to disclose. Manuscript received December 17, 2012; revised manuscript received January 15, 2014, accepted January 17, 2014.
patients with CHF (6–8). Furthermore, there is little evidence that lower resting ventricular rate in patients with CHF and AF is important physiologically or improves quality of life (9–11). Accordingly, we investigated the relationship between ventricular rate and survival in patients with CHF due to LVSD who were in AF or SR. Methods Study population and data collection. We included outpatients with CHF who attended a community-based heart failure program in Kingston-Upon-Hull, United Kingdom, serving a population of 600,000 people, between September 1999 and October 2010. Patients were referred for diagnosis and/or management of CHF. Patients gave written consent for their data to be stored electronically and used for research purposes. A large proportion of surviving patients were reassessed at 1 year after attempts had
2
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
been made to optimize therapy in accordance with UK guidelines for the management of ACE-I = angiotensinchronic heart failure (12). converting enzyme inhibitor All patients underwent a AF = atrial fibrillation standardized clinical examinaARB = angiotensin receptor tion, 12-lead electrocardiography, blocker echocardiography, spirometry, CHF = chronic heart failure and routine blood tests. Cardiac DBP = diastolic blood rhythm and resting heart rate pressure were determined from a 12-lead HR = heart rate electrocardiogram, which was IHD = ischemic heart obtained after at least 5 min rest in disease the supine position, using a MAC IQR = interquartile range 5000 machine (GE Healthcare, LBBB = left bundle branch Milwaukee, Wisconsin). We used block the arbitrary definitions of the LVSD = left ventricular RACE II (Rate Control Efficacy systolic dysfunction in Permanent Atrial Fibrillation: NYHA = New York Heart a Comparison between Lenient Association versus Strict Rate Control II) SBP = systolic blood study to categorize patients in AF pressure into those with “lenient” (80 SR = sinus rhythm and <110 per min) or “strict” (<80 per min) ventricular rate control (7). Patients who were permanently paced were excluded from the analysis. All echocardiograms were performed by experienced echocardiographers in accordance with recommendations of the British Society of Echocardiography. For echocardiogram acquisition, a Vivid 5 or Vivid 7 system (GE Healthcare) using a 3.4-MHz probe was used. When possible, the Simpson’s biplane method was used to measure the left ventricular ejection fraction (LVEF), and, in all cases, a visual approximation was made. Only patients with at least mild left ventricular systolic dysfunction (LVSD) were included. The primary endpoint of the study was all-cause mortality. Mortality data were captured electronically by National Health Service information systems and reported to the hospital and were complete to the censor date of August 31, 2011. Ethics. The investigation conformed to the principles outlined in the Declaration of Helsinki. It was approved by the Hull and East Yorkshire Research Ethics Committee (Heart Care Study ELSY 2642). All subjects gave written informed consent. Statistical methods. Data were tested for normality using the Kolmogorov-Smirnov test. Normally distributed data are presented as mean SD; other data are presented as medians and interquartile ranges (IQR). Categorical data are given as percents. Differences in categorical variables between groups were compared with the chi-square test. Differences between groups with normally distributed data were analyzed using Student’s t test, and non-normally distributed data were analyzed using the Mann-Whitney U test. Abbreviations and Acronyms
Kaplan-Meier curves used for analysis of survival and significance tests between groups were tested using the logrank test. Univariate and multivariate Cox regression models were constructed to investigate the relationship between potential covariates and outcome. To avoid overfitting, we included only variables significantly associated with outcome in the univariate analysis (p < 0.05) in the forward conditional stepwise Cox analysis. Proportional hazards assumptions were checked by Schoenfeld residuals, and Martingale residuals plots were used to evaluate linearity (13,14). Results are presented as hazard ratios (HRs) with 95% confidence intervals (CIs). Analyses were performed using SPSS version 19 (IBM Corp., Armonk, New York) and Stata 13 software (StataCorp., College Station, Texas). Results Baseline patient characteristics. A total of 2,039 patients was included in the baseline analysis, of whom 24% (n ¼ 488) had AF (Table 1). Compared to patients in SR, patients in AF were older and less likely to be diabetic or to have ischemic heart disease. Although there were no differences in severity of LVSD, patients with AF were more symptomatic than patients in SR. Patients with AF were less likely to be prescribed a beta-blocker, aldosterone antagonist, statin, or aspirin but more likely to be prescribed warfarin, digoxin, or a loop diuretic agent. The median resting heart rate was higher for patients with AF than for patients with SR (81 per min [IQR: 69 to 98] versus 70 (IQR: 60 to 83) per min, respectively, p ¼ 0.0001). Relationship between heart rhythm and survival. During the follow-up period, 229 patients (47%) with AF died compared to 639 patients (41%) with SR. The 1-year mortality rate for patients with AF was 15% (n ¼ 74) versus 11% for patients with SR (n ¼ 170). Overall median follow-up was 3.6 years (IQR: 1.7 to 6.9 years). The overall median survival for patients with AF was 6.1 years (IQR: 5.3 to 6.9 years) compared to 7.3 years (IQR: 6.5 to 8.1 years) for patients with SR. In a multivariable model using baseline data that was corrected for age, sex, weight, QRS duration, heart rate, systolic blood pressure, New York Heart Association (NYHA) class III/IV versus I/II, ischemic heart disease (IHD) status (yes/no), diabetic status (yes/no), angiotensinconverting enzyme/angiotensin receptor blocker (ACE/ ARB) use (yes/no), LVSD severity (moderate vs.
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Table 1
3
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
Characteristics of Patients in AF and SR at Baseline and at 1-Year Follow-Up Baseline Review % of Patients in AF (n ¼ 488)
Age, yrs Male Weight, kg
% of Total Patient Cohort
“Strict” HR (n ¼ 219)
“Lenient” HR (n ¼ 200)
HR 110 beats/min (n ¼ 69)
76 (70–81)
74 (68–80)
72 (63–78)
75
76
80 (67–92)
83 (69–97)
Patients in AF (n ¼ 488)
Patients In SR (n ¼ 1,551)
p Value
0.01
74 (68–80)
70 (63–77)
0.0001
64
0.11
74
72
0.45
84 (69–97)
0.25
81.6 (68–95)
78.9 (67–91)
0.01
0.04
15
19
0.03
0.0001
53
69
Diabetic patient
15
19
6
IHD
61
52
32
NYHA functional class
p Value
I
11
7
9
9
17
II
53
45
45
49
51
III
35
43
41
40
29
IV
1
3
6
2
2
SBP, mm Hg
130 (112–147)
127 (116–146)
122 (110–136)
128 (112–145)
130 (115–147)
DBP, mm Hg
74 (65–87)
80 (70–90)
84 (76–96)
79 (68–90)
75 (66–83)
Mild
23
17
16
19
17
Mild-to-moderate
14
18
13
15
12
Moderate
25
18
7
20
23
Moderate-to-severe
21
22
26
22
23
Severe
17
26
38
24
25
67 (59–73)
90 (85–98)
121 (113–132)
0.0001
81 (69–98)
70 (60–83)
112 (98–138)
104 (92–120)
102 (90–116)
0.0001
106 (96–128)
110 (96–136)
HR, beats/min (IQR)
0.0001 0.003
LVSD
QRS duration, ms (IQR)
0.12
0.0001 0.10 0.02
None
39
51
67
48
Low dose
29
27
22
32
34
Moderate dose
24
18
12
15
19
High dose
0.09 0.0001 0.18
0.002
Beta-blocker*
0.0001 0.0001
0.09
40
7
5
0
5
7
ACE-I/ARB
82
68
77
0.003
76
79
0.07
Aldosterone antagonist
25
20
7
0.005
20
25
0.03
Loop diuretic agent
83
77
81
0.29
80
71
0.0001
Digoxin
51
51
48
0.88
51
8
0.0001
Amiodarone
10
8
7
0.69
8
8
0.70
1
5
7
0.03
3
3
0.72
Statin
41
35
23
0.03
36
57
0.0001
Warfarin
38
49
13
0.06
58
12
0.0001
Aspirin
50
38
13
0.06
29
59
0.0001
5
5
3
0.75
13
5
0.0001
Verapamil/diltiazem
Clopidogrel
Continued on the next page
with a 7% decrease in mortality per year (HR: 0.93; 95% CI: 0.88 to 0.99, p ¼ 0.02) (Table 2). After multivariable adjustment, heart rate as a continuous variable was not significantly associated with survival (HR: 0.94; 95% CI: 0.88 to 1.00, p ¼ 0.07), nor was it when split by quartiles (Fig. 1). For patients in SR, in the multivariable model, an increase in resting heart rate by 10 beats/min was associated with a 10% increase in mortality per year (HR: 1.10; CI: 95% 1.04 to 1.15, p <0.0001]. The group was divided into quartiles according to heart rate: <60/min (referent); 60 to 70/min; 71 to 83/min, and >83/min. Compared to the referent group, only patients with a resting heart rate >83/min had
significantly worse survival (HR: 1.69; CI: 1.30 to 2.21, p ¼ 0.0001). Patient characteristics at 1 year. A total of 841 patients was seen after the first year of active treatment in the clinic, of whom 22% (n ¼ 184) were in AF (Table 1). Patients who had AF were older and weighed more but otherwise had similar characteristics to patients in SR. Patients with AF had higher resting heart rate than patients in SR (72 beats/min [IQR: 63 to 82]) vs. 63 beats/min [IQR: 57 to 71], respectively; p ¼ 0.0001). Prescription of betablockers, ACE inhibitors/ARBs, or aldosterone antagonists was similar in the 2 groups. Patients in AF were more likely to be prescribed loop diuretic agents, digoxin, and
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
4
Table 1
Continued 1- Year Review % of Patients in AF (n ¼ 184)
% of Total Patient Cohort
“Strict” HR (n ¼ 129)
“Lenient” HR (n ¼ 48)
HR 110 beats/min (n ¼ 7)
73 (67–79)
75 (64–80)
77 (66–79)
82
79
85 (74–97)
Patients in AF (n ¼ 184)
Patients in SR (n ¼ 657)
p Value
0.85
73 (66–79)
69 (62–75)
0.0001
57
0.26
80
75
0.14
83 (69–96)
73 (69–100)
0.43
84.6 (73–97)
81.1 (70–92)
0.03
19
28
17
0.45
20
18
0.38
66
60
57
0.74
69
64
p Value
0.01
0.25 0.06
12
13
43
13
22
71
52
14
64
60
16
33
43
22
17
2
2
0
2
1
124 (110–141)
127 (112–151)
116 (103–134)
0.37
125 (110–141)
126 (115–143)
0.16
72 (65–83)
80 (68–92)
65 (62–86)
0.06
73 (65–84)
72 (64–80)
0.08
24
29
14
25
22
15
6
43
14
14
19
23
14
20
23
27
17
14
24
24
15
25
14
17
17
65 (59–73)
85 (83–91)
120 (114–126)
0.0001
72 (63–82)
63 (57–71)
114 (102–144)
110 (94–134)
90 (84–96)
0.0001
114 (98–142)
110 (96–138)
0.17
0.91
0.005
0.0001 0.44 0.75
11
27
29
16
14
41
19
57
36
34
33
25
14
30
34
16
29
0
18
18
94
94
100
0.79
94
92
0.31
35
23
29
0.31
32
28
0.35
88
85
100
0.58
88
74
0.0001
55
52
43
0.79
54
10
0.0001
12
10
0
0.58
11
10
0.47
1
4
0
0.27
2
2
0.86
43
31
57
0.25
40
66
0.0001
20
23
29
0.82
78
18
0.0001
83
67
71
0.06
21
55
0.0001
4
0
14
0.11
3
12
0.0001
Values are median (interquartile range) or n. *Doses of beta-blocker: None, 0%; Low dose, 1% to 33%; Moderate dose, 34% to 66%; High dose, 67% to 100%. Values of p 0.05 are in bold. ACE-I/ARB ¼ angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; AF ¼ atrial fibrillation; DBP ¼ diastolic blood pressure; HR ¼ heart rate; IHD ¼ ischemic heart disease; IQR ¼ interquartile range; LBBB ¼ left bundle branch block; LVSD ¼ left ventricular systolic dysfunction; NYHA ¼ New York heart association; SBP ¼ systolic blood pressure; SR ¼ sinus rhythm.
warfarin but less likely to be prescribed statins, aspirin, or clopidogrel. Relationship between heart rate at 1 year and survival. In univariable analysis restricted to the patients in AF, ventricular rate was not a significant predictor of survival (HR: per 10 beats/min increase in heart rate, 1.01; 95% CI: 0.90 to 1.13, p ¼ 0.84) (Fig. 2; Table 3). In contrast, for patients in SR, heart rate was significantly associated with survival. In the multivariable model, each increase in resting heart rate by 10 beats/min was associated with an 13% increase in annual mortality (HR: 1.13, 95%
CI: 1.03 to 1.24, p ¼ 0.008). The group was subdivided into quartiles according to heart rate: <57 beats/min; 57 to 62 beats/min; 63 to 70 beats/min; and >70 beats/min. Only patients with a resting heart rate >70 beats/min had worse survival than referent group (<57 beats/min; HR: 1.47; 95% CI: 1.03 to 2.13, p ¼ 0.04). Discussion This analysis suggests that after adjusting for age and other differences, patients with CHF and LVSD who are
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Table 2
Univariable and Multivariable Analyses of Patients in AF and SR at Baseline* Patients in AF HR Represents
Age
Per decade increase
Patients in SR
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
1.65 (1.41–1.93)
<0.0001
1.53 (1.29–1.82)
<0.0001
1.76 (1.61–1.92)
<0.0001
1.73 (1.58–1.90)
<0.0001
Sex
Men vs. women
1.09 (0.81–1.45)
0.59
–
–
1.26 (1.06–1.49)
0.008
–
–
Weight
Per 5-kg increase
0.95 (0.92–0.98)
0.004
–
–
0.92 (0.90–0.94)
<0.0001
–
–
1.05 (1.005–1.10)
QRS duration
Per 10-ms increase
1.07 (1.02–1.12)
0.004
1.03 (1.00–1.05)
0.04
–
–
LVI severity
Moderate vs.
1.30 (0.97–1.73)
0.08
–
–
1.30 (1.08–1.56)
0.005
–
–
Heart rate
Per 10-beats/min increase
0.93 (0.88–0.99)
0.02
–
–
1.15 (1.10–1.20)
<0.0001
1.10 (1.05–1.15)
<0.0001
Systolic BP
Per 5-U increase
0.97 (0.94–1.00)
0.03
–
–
0.97 (0.96–0.99)
0.001
0.97 (0.96–0.99)
0.002
NYHA class
III/IV vs. I/II
1.82 (1.40–2.38)
<0.0001
1.51 (1.14–1.99)
0.004
1.80 (1.53–2.12)
<0.0001
1.32 (1.11–1.57)
0.002
IHD
Yes vs. no
1.63 (1.25–2.13)
<0.0001
1.53 (1.14–2.04)
0.005
1.09 (0.92–1.30)
0.32
Diabetic patient
Yes vs. no
1.15 (0.80–1.64)
–
–
1.31 (1.09–1.59)
0.005
0.45
0.03
–
–
1.36 (1.11–1.66)
0.003
Yes vs. no
0.86 (0.65–1.15)
0.32
–
–
0.78 (0.65–0.93)
0.006
0.80 (0.66–0.97)
0.021
Yes vs. no
1.34 (0.94–1.89)
0.10
–
–
2.21 (1.81–2.70)
<0.0001
1.69 (1.35–2.11)
<0.0001 –
0.22
–
Aldosterone Antagonist
Yes vs. no
1.22 (0.89–1.69)
Beta-blocker
Yes vs. no
0.67 (0.516–0.874)
Digoxin
Yes vs. no
0.92 (0.71–1.19)
0.50
–
Amiodarone
Yes vs. no
1.27 (0.83–1.93)
0.28
–
0.003
0.71 (0.53–0.94)
–
1.29 (1.08–1.53)
0.005
–
0.71 (0.60–0.82)
<0.0001
–
–
–
1.48 (1.14–1.93)
0.003
1.50 (1.14–1.98)
0.004
–
1.31 (1.01–1.70)
0.046
1.37 (1.04–1.80)
0.026
0.02
*Hazard ratios for primary endpoint mortality. yVariables retained in the multivariate Cox model are shown. Values in bold indicate hazard ratios with a value of p 0.05. ACE/ARB ¼ angiotensin-converting enzyme/angiotensin blocker receptor; AF ¼ atrial fibrillation; BP ¼ blood pressure; CI ¼ confidence interval; HR ¼ hazard ratio; IHD ¼ ischemic heart disease; LVSD ¼ left ventricular systolic dysfunction; MVM ¼ multivariate model; NYHA ¼ New York Heart Association; SR ¼ sinus rhythm; UVM ¼ univariate model.
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
ACE/ARB Loop diuretic agent use
5
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
6
Figure 1
Multivariable Adjusted Survival Curves for Patients in Atrial Fibrillation at Baseline Divided by Heart Rate Quartiles
CI ¼ confidence interval; HR ¼ hazard ratio.
in AF or SR have a similar prognosis, despite patients in AF having a substantially higher ventricular rate. After 1 year of follow-up and optimization of medical therapy, resting ventricular rate was <80 beats/min in 70% of patients with AF and the median ventricular rate was only 72 beats/min (IQR: 63 to 83 beats/min), although this remained somewhat higher than the median rate for patients in SR (63 beats/min; IQR: 57 to 71 beats/min). Thus, adequate rate control might have contributed to the similar survival of patients in AF and SR. In contrast to patients in SR, higher ventricular rates were not associated with increased mortality in patients with CHF, LVSD, and AF. Similar conclusions have been found from post-hoc analysis of the CHARM (Candesartan in Heart Failure: Assessment of Reduction in Mortality and morbidity) studies (15). There is a strong association between ventricular rate and mortality for patients with CHF in SR (4). If this is solely a rate-mediated effect, then a similar relationship would be expected for patients in AF. However, a growing body of evidence suggests that ventricular rate may be less important among patients with AF (8,9,11). Pooled data from the AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) and
Figure 2
Multivariable Adjusted Survival Curves for Patients in Atrial Fibrillation after 1-Year Follow-Up Divided by Heart Rate Quartiles
Abbreviations as in Figure 1.
RACE studies suggest that chronic resting heart rates >100 beats/min might be detrimental (16). Current recommendations suggest that resting ventricular rate should be lowered to <100 beats/min for the asymptomatic patient and, more aggressively (i.e., <80 beats/min at rest and <110 beats/min with moderate exercise) if symptoms persist or if LVEF deteriorates (17,18). Some patients with AF and CHF have better symptom control and higher LVEF when a lower ventricular rate is achieved by using a combination of carvedilol and digoxin (10). The RACE II study (n ¼ 614), is the only prospectively designed randomized controlled trial to investigate whether “strict” or “lenient” ventricular rate control should be preferred in patients with AF (7). Neither strategy was associated with any significant difference in mortality, hospitalizations, or symptoms during 3-year follow-up. Only 10% of the patients enrolled had a history of heart failure, and the mean LVEF was 51%, thus, extrapolation of the data to steer the management of patients with CHF would seem inappropriate. Clearly, there must be a ventricular rate below which harm occurs. The notion that sudden cardiac death in patients with CHF is largely due to tachyarrhythmia may be exaggerated. The CARISMA (Cardiac Arrhythmias
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Table 3
Univariable and Multivariable Analyses of Patients in AF and SR at 1 Year* Patients in AF HR Represents
Patients in SR
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
UVM HR (95% CI)
p Value
MVM HR (95% CI)y
p Value
1.55 (1.21–1.99)
0.001
2.16 (1.88–2.50)
<0.0001
2.01 (1.75–2.33)
<0.0001
Age
Per decade increase
1.51 (1.19–1.91)
0.001
Sex
Men vs. women
1.004 (0.62–1.62)
0.99
–
–
1.25 (0.96–1.63)
–
–
Weight
Per 5-kg increase
0.96 (0.91–1.01)
0.15
–
–
0.91 (0.88–0.95)
<0.0001
–
– 0.035
0.09
Per 10-ms increase
1.05 (0.99–1.13)
0.11
–
–
1.05 (1.03–1.08)
<0.0001
1.03 (1.00–1.06)
Moderate vs.
1.65 (1.09–2.49)
0.02
1.61 (1.07–2.45)
0.024
1.56 (1.21–2.02)
0.001
1.36 (1.04–0.78)
0.025
Heart rate
Per 10-beats/min increase
1.01 (0.90–1.13)
0.84
–
–
1.13 (1.03–1.23)
0.008
1.13 (1.03–1.24)
0.008
Systolic BP
Per 5-U increase
1.00 (0.96–1.04)
0.95
–
–
1.00 (0.97–1.02)
NYHA class
III/IV vs. I/II
1.53 (0.99–2.35)
0.05
–
–
2.26 (1.73–2.95)
IHD
Yes vs. no
1.10 (0.73–1.65)
0.66
–
–
0.96 (0.75–1.24)
Diabetic patient
Yes vs. no
1.41 (0.89–2.23)
0.15
–
–
1.30 (0.97–1.74)
ACE/ARB use
Yes vs. no
0,96 (0.42–2.20)
0.93
–
–
0.67 (0.46–0.98)
Loop diuretic use
Yes vs. no
2.47 (1.15–5.33)
0.02
2.48 (1.15–5.36)
0.021
2.10 (1.53–2.87)
Aldosterone antagonist use
Yes vs. no
0.90 (0.58–1.38)
0.61
–
–
1.20 (0.93–1.55)
Beta-blocker use
Yes vs. no
0.83 (0.49–1.40)
0.48
–
–
0.53 (0.40–0.71)
–
–
1.68 (1.26–2.25)
<0.0001
0.76
–
–
0.08
–
–
0.04
–
–
1.42 (1.02–1.99)
0.039
0.77 <0.0001
<0.0001
–
–
<0.0001
–
–
0.17
Digoxin use
Yes vs. no
0.91 (0.62–1.35)
0.65
–
–
1.74 (1.17–2.57)
0.006
–
–
Amiodarone use
Yes vs. no
1.24 (0.69–2.22)
0.48
–
–
1.13 (0.77–1.66)
0.54
–
–
*Hazard ratios for the primary endpoint mortality. yVariables retained in the multivariate Cox model are shown. Values in bold indicate hazard ratios with a value of p 0.05. Abbreviations as in Table 2.
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
QRS duration LVSD severity
7
8
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
and Risk Stratification After Acute Myocardial Infarction) trial analyzed data over 2 years from implantable loop recorders in 297 patients with IHD and a reduced LVEF (19). After multivariate adjustment, high degree atrioventricular block was the strongest predictor of cardiac death (HR: 6.75, 95% CI: 2.55 to 17.84, p < 0.001), followed by sinus bradycardia (HR: 4.15, 95% CI: 1.37 to 12.62, p ¼ 0.012). The number of patients enrolled with AF was small (n ¼ 8 permanent and n ¼ 17 paroxysmal), but half of the patients with permanent, and 12% with paroxysmal AF went on to develop high-degree atrioventricular block compared to only 8% of patients in SR (20). Similarly, patients in strict heart rate control groups in the AFFIRM and RACE trials (most of whom did not have LVSD) had a higher incidence of pacemaker implantation (16). In a substudy of PRIME II (Second Prospective Randomised Study of Ibopamine on Mortality and Efficacy), patients with AF and advanced heart failure with a baseline resting heart rate <80 beats/min had worse survival (HR: 2.9, 95% CI: 1.4 to 5.8, p ¼ 0.002) than patients with heart rates >80 beats/min (21). Atrial fibrillation causes loss of “atrial kick,” reduced LV diastolic filling, and a fall in stroke volume. An increase in ventricular rate may compensate for these changes to maintain adequate cardiac output; therefore, aggressive rate lowering in patients with CHF and AF may be hemodynamically detrimental (22). Tight daytime control of ventricular rate may, for selected patients, be associated with an exacerbation of nocturnal pauses which may increase the likelihood of pause-dependent ventricular tachycardia (10,23). The prevalence of AF in patients with CHF is estimated to be 30% to 40%, which increases with worsening LVSD (24). The baseline prevalence of AF in randomized trials of betablockers to treat CHF ranges from 13% to 35%. Bisoprolol, metoprolol, and nebivolol have not been shown to improve outcomes in patients with AF and CHF, although carvedilol may (25,26). If beta-blockers are less effective or ineffective in patients with heart failure and AF, this may reflect the lack of importance of ventricular rate control or that exacerbation of nocturnal pauses detracts from the benefits of rate control. Settling this issue is important for future clinical practice. Study limitations. This was a post-hoc analysis of prospectively collected data with all the limitations that imposes. Resting heart rate is a simple but poor indicator of heart rate control during activity, and we appreciate that a better assessment of risk might be obtained from 24-h monitoring. The analysis does not control for any medical treatment changes or pacemaker implantation after 1 year of follow-up. We sampled patients at 2 time points and did not control for patients who switched rhythms, although such patients were few. Atrial fibrillation may be an independent predictor of an adverse outcome in patients with CHF and preserved ejection fraction, but we did not explore this because follow-up data collection was restricted to patients with LVSD. We could not define whether beta-blockers were associated with a
lower mortality in our observational study, but the analysis may have lacked power and could not adjust for unmeasured confounders. Conclusions For patients with chronic heart failure in SR, lower heart rates are associated with lower mortality, but this is not the case for similar patients in AF. Whether it is advantageous to lower heart rate in CHF patients with AF to a range similar to that recommended for patients in sinus rhythm (<70 beats/min) is uncertain and requires further investigation. The benefits of daytime ventricular rate reduction may be balanced by an increased risk, possibly due to exacerbation of nocturnal pauses.
Reprint requests and correspondence: Dr. Damien Cullington, Yorkshire Heart Centre, Department of Adult Congenital Heart Disease, Jubilee Building, Leeds General Infirmary, Leeds, West Yorkshire LS1 3EX, United Kingdom. E-mail: [email protected].
REFERENCES
1. Diaz A, Bourassa MG, Guertin MC, Tardif JC. Long-term prognostic value of resting heart rate in patients with suspected or proven coronary artery disease. Eur Heart J 2005;26:967–74. 2. Palatini P, Thijs L, Staessen JA, et al., for the Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Predictive value of clinic and ambulatory heart rate for mortlality in elderly subjects with systolic hypertension. Arch Intern Med 2002;162:2313–21. 3. Dyer AR, Persky V, Stamler J, et al. Heart rate as a prognostic factor for coronary heart disease and mortality : findings in three Chicago epidemiological studies. Am J Epidemiol 1980;112:737–49. 4. Böhm M, Swedberg K, Komajda M, Borer JS. Heart rate as a risk factor in chronic heart failure (SHIFT): the association between heart rate and outcomes in a randomised placebo-controlled trial. Lancet 2010;376:886–94. 5. Böhm M, Borer J, Ford I, et al. Heart rate at baseline influences the effect of ivabradine on cardiovascular outcomes in chronic heart failure: analysis from the SHIFT study. Clin Res Cardiol 2013;102:11–22. 6. Heist EK, Mansour M, Ruskin JN. Rate control in atrial fibrillation. Circulation 2011;124:2746–55. 7. Van Gelder IC, Groenveld HF, Crijns HJ, et al., for the RACE II Investigators. Lenient versus strict rate control in patients with atrial fibrillation. N Engl J Med 2010;362:1363–73. 8. Cooper HA, Bloomfield DA, Bush DE, et al., for the AFFIRM Investigators. Relation between achieved heart rate and outcomes in patients with atrial fibrillation (from the Atrial Fibrillation Follow-up Investigation of Rhythm Management [AFFIRM] Study). Am J Cardiol 2004;93:1247–53. 9. Groenveld HF, Crijns HJ, Van den Berg MP, et al., for the RACE II Investigators. The effect of rate control on quality of life in patients with permanent atrial fibrillation. J Am Coll Cardiol 2011;58: 1798–803. 10. Khand AU, Rankin AC, Martin W, Taylor J, Gemmell I, Cleland JGF. Carvedilol alone or in combination with digoxin for the management of atrial fibrillation in patients with heart failure. J Am Coll Cardiol 2003;42:1944–51. 11. Smit MD, Crijns HJ, Tijssen JG, et al., for the RACE II Investigators. Effect of lenient versus strict rate control on cardiac remodelling in patients with atrial fibrillation. J Am Coll Cardiol 2011;58:942–9. 12. National Institute for Health and Care Institute. Chronic Heart Failure. National Clinical Guideline for Diagnosis and Management in
JACC: Heart Failure Vol. -, No. -, 2014 - 2014:-–-
13. 14. 15.
16.
17.
18. 19.
Primary and Secondary Care. Available at: www.nice.org.uk/CG108. Accessed June 2012. Grambsch PM, Therneau TM. Proportional hazards and diagnostics based on weighted residuals. Biometrika 1994;81:515–26. Therneau TM, Grambsch PM, Fleming TR. Martingale based residuals for survival models. Biometrika 1990;77:147–60. Castagno D, Skali H, Takeuchi M, et al., for the CHARM Investigators. Association of heart rate and outcomes in a broad spectrum of patients with chronic heart failure: results from the CHARM (Candesartan in Heart Failure: Assessment of Reduction in Mortality and morbidity) program. J Am Coll Cardiol 2012;59: 1785–95. Van Gelder IC, Wyse DG, Chandler ML, et al., for the RACE and AFFIRM Investigators. Does intensity of rate-conrtol influence outcome in atrial fibrillation? An analysis of pooled data from the RACE and AFFIRM studies. Europace 2006;8:935–42. Gillis AM, Verma A, Talajic M, Nattel S, Dorian Pfor the CCS Atrial Fibrillation Guidelines Committee. Canadian Cardiovascular Society atrial fibrillation guidelines 2010: rate and rhythm management. Can J Cardiol 2011;27:47–59. Wasmer K, Eckardt L. Management of atrial fibrillation around the world: a comparison of current ACCF/AHA/HRS/CCS, and ESC guidelines. Europace 2011;13:1368–74. Bloch Thomsen PE, Jons C, Raatikainen MJ, et al., for the Cardiac Arrhythmias and Risk Stratification After Acute Myocardial Infarction (CARISMA) Study Group. Long-term recording of cardiac arrhythmias with an implantable cardiac monitor in patients with reduced ejection fraction after acute myocardial infarction. The Cardiac Arrhythmias and Risk Stratification After Acute Myocardial Infaction (CARISMA) Study. Circulation 2010;121:1258–64.
9
Cullington et al. Importance of Heart Rate in Patients With CHF and AF
20. Gang UJ, Jøns C, Jørgensen RM, et al., for the CARISMA investigators. Risk markers of late high-degree atrioventricular block in patients with left ventricular dysfuction after an acute myocardial infarction: a CARISMA substudy. Europace 2011;13:1471–7. 21. Rienstra M, Van Gelder IC, Van den Berg MP, Boomsma F, Hillege HL, Van Veldhuisen DJ. A comparison of low versus high heart rate in patients with atrial fibrillation and advanced chronic heart failure: effects on clinical profile, neurohormonaes and survival. Int J Cardiol 2006;109:95–100. 22. Rawles JM. What is meant by a “controlled” ventricular rate in atrial fibrillation? Br Heart J 1990;63:157–61. 23. Sweeney MO, Ruetz LL, Belk P, Mullen TJ, Johnson JW, Sheldon T. Bradycardia pacing-induced short-long-short sequences at the onset of ventricular tachyarrhythmias: a possible mechanism of proarrhythmia? J Am Coll Cardiol 2007;50:614–22. 24. Maggioni AP, Dahlström U, Filippatos G, et al., for the Heart Failure Association of European Society of Cardiology. EURObservational research program: the Heart Failure Pilot Survey (ESC-HF Pilot). Eur J Heart Fail 2010;12:1076–84. 25. Joglar JA, Acusta AP, Shusterman NH, et al. Effect of carvedilol on survival and haemodynamics in patients with atrial fibrillation and left ventricular dysfunction: retrospective analysis of the US carvedilol heart failure trials program. Am Heart J 2001;142:498–501. 26. Rienstra M, Damman K, Mulder BA, Van Gelder IC, McMurray JJV, Van Veldhuisen DJ. Beta-blockers and outcome in heart failure and atrial fibrillation: a meta-analysis. J Am Coll Cardiol HF 2013;1:21–8. Key Words: atrial fibrillation sinus rhythm.
-
heart failure
-
heart rate
-
prognosis
-