Author + information
- Michael R. Bristow, MD, PhD∗ ( and )
- Natasha L. Altman, MD
- ↵∗Address for correspondence:
Dr. Michael R. Bristow, University of Colorado Division of Cardiology and Cardiovascular Institute, 12700 East 19th Avenue, Campus Box B139, Aurora, Colorado 80045.
Heart rate has assumed a prominent role in the pathophysiology and natural history of heart failure (HF). In reduced left ventricular ejection fraction (EFrHF), higher heart rates are associated with increased morbidity and mortality (1). Rapid pacing is one of the most common animal models of systolic dysfunction heart failure, and persistent tachycardia may produce a cardiomyopathy in humans (2). The most direct evidence that higher heart rates that are within or only slightly above the normal range have adverse effects on HFrEF major clinical outcomes derives from SHIFT (Systolic Heart Failure Treatment With the IF inhibitor ivabradine Trial) trial (3), where the heart rate-lowering agent ivabradine produced a reduction of 18% (p < 0.0001) in the primary endpoint of cardiovascular (CV) mortality or HF hospitalization. In addition, part of the favorable mechanism of action of β-blockers in sinus rhythm HFrEF patients is almost certainly heart rate reduction (4,5).
What about heart rate in preserved ejection fraction HF (HFpEF)? A previous post-hoc analysis from the TOPCAT (Treatment Of Preserved Cardiac function HF with an Aldosterone anTagonist) trial database confined to the 2,705 patients in sinus rhythm found that increasingly higher heart rates measured at baseline, pre-randomization electrocardiography were directly related to the hazards for all-cause or HF hospitalization and all-cause or CV mortality (6). In this issue of JACC: Heart Failure, Vazir et al. (7), based on another post hoc analysis of TOPCAT in patients in any rhythm, report that higher baseline, most recently measured or within-trial change (Δ) from the most recently measured heart rate, was associated with an elevated hazard for the trial’s primary endpoint of time to CV death, HF hospitalization, or aborted cardiac arrest. The increase in hazard was small, 8% to 11%, but highly statistically significant (p < 0.001) in the primary adjusted model (7). The primary analysis was conducted in the combined treatment groups of the “Americas” TOPCAT cohort of 1,767 patients, in order “to have a more homogenous group of subjects with contemporary definition of HFpEF” (7). However, the cohort consisting of Russians and Georgians demonstrated qualitatively similar but even larger heart rate effects on the primary composite and its 2 major components (7). There was no evidence of an increased risk of myocardial infarction with higher or increasing heart rates, providing support for the HF specificity of the findings. On the other hand, noncardiac deaths, adjudicated by an endpoints committee in TOPCAT (Online Ref. 1) exhibited an even greater hazard with higher or Δ heart rate increases, 12% to 20% (p = 0.002 and p < 0.001, respectively). This finding was interpreted by the authors to mean that higher or increasing heart rates are markers for “….a non-specific signal of deteriorating health or episodes of acute infection or other systemic stress” (7). This interpretation was further supported by a decrease in non-CV deaths but not HF outcomes associated with reductions in heart rate (7). These observations differ from those from the CHARM (Candesartan in Heart failure Assessment of Reduction in Mortality and morbidity) trial, where a decrease in heart rate was associated with a reduced hazard for CV death, HF hospitalization, or its composite and a statistically significant reduction in non-CV death hazard (Online Ref. 2).
The multivariate analyses in Vazir et al. (7) accounted for spironolactone treatment effects (in a second model), beta-blocker use and dose, presence of atrial fibrillation, and other variables that affected outcomes in a univariate analysis. It would have been helpful for the supplementary material to have included a within-treatment group analysis, in view of the significant reduction that spironolactone produced in the primary endpoint hazard ratio in the TOPCAT Americas cohort (Online Ref. 1). In addition, a more ideal study design would have been to prospectively designate heart rate effects on outcomes as an objective in the protocol or statistical analysis plan. This presumably would have led to a standardized, more controlled method of measuring heart rate in the trial. In that sense, it is likely the reported findings are an underestimation of the true effect of heart rate on outcomes. Although the findings need to be confirmed in another HFpEF cohort, the consistency of the findings by Vazir et al. (7), the separate TOPCAT baseline heart rate analysis in sinus rhythm patients (6), and the support from other sources argue for validity of the data and the authors’ interpretations.
The recently reported EDIFY (prEserveD left ventricular ejectIon fraction chronic heart Failure with ivabradine study) findings of ivabradine in HFpEF subjects (Online Ref. 3) also generally support the findings of Vazir et al. (7). The EDIFY investigators randomized 179 HFpEF patients with heart rates ≥70 beats/min (the same entry criterion as in SHIFT ) and left ventricular EFs (LVEFs) of ≥0.45 (same as TOPCAT) (Online Ref. 1) to receive ivabradine or placebo, with 3 coprimary endpoints of E/e′ by using echo-Doppler, the 6-min walk test, and plasma N-terminal pro–B-type natriuretic peptide concentration (Online Ref. 3). Ivabradine decreased placebo heart rate of 9.5 beats/min, into the low 60s, but there were no changes and no trends in any of the primary endpoints (Online Ref. 3). Collectively, the EDIFY data (Online Ref. 3) and the observation by Vasir et al. (7) that a reduction in heart rate in TOPCAT was not associated with a reduced hazard for HF endpoints likely means there is little chance that therapeutically lowering heart rate will be of benefit in the general HFpEF patient population. Monitoring of heart rate in HFpEF and reacting to an increase by intensifying HF therapy or instituting a diagnostic search for an emerging or worsening comorbidity may be helpful, but heart rate is only 1 of many clinical clues that would be available to an astute clinician in these situations. In addition, for detection of worsening HF by clinical heart rate monitoring, an implantable, insertable, or wearable cardiac monitor with remote detection capability would likely be much more effective than brief duration snapshot clinical heart rates measured by multiple methods. In addition, in HF, heart rate may not increase when pulmonary pressures rise to levels that require urgent diuresis (Online Ref. 4).
The collective interpretation of the CHARM (Online Ref. 2), SHIFT (3), TOPCAT (Online Ref. 1), and EDIFY (Online Ref. 3) trial data lead to the conclusion that heart rate lowering reduces major HF clinical events in HFrEF but not in HFpEF. Why should this be, if both phenotypes are associated with the clinical syndrome of HF and both exhibit increased heart rates due at least in part to heightened cardiac adrenergic activity? One obvious possibility is that an increase in heart rate is a major means of compensation when the left ventricle develops any type of impaired performance and, because HFpEF ventricles exhibiting diastolic dysfunction do not have the ability to stabilize cardiac output by increasing stroke volume through increasing end diastolic volume, they are more heart-rate dependent. When a HFrEF left ventricle undergoes heart rate lowering by a beta-blocker, it eventually exhibits increased intrinsic contractile function due at least in part to favorable changes in the expression of contractility-modulating genes that are regulated by beta1-adrenergic receptor signaling (Online Refs. 5,6). The improved contractile function compensates for the heart rate reduction effect on cardiac output and ultimately leads to reverse remodeling and a decrease in end diastolic volume (Online Ref. 7). In HFrEF, heart rate lowering with ivabradine appears to produce a qualitatively similar but lesser degree of reverse remodeling (Online Ref. 8), and the role of heart rate lowering per se on gene expression is under investigation (Online Ref. 9). Therefore, in the absence of a therapeutic means to improve diastolic function and ventricular filling, the HFpEF left ventricle likely needs to operate at a higher heart rate as a compensatory mechanism, and lowering heart rate may not be beneficial.
Finally, the heart rate behavior story in HFrEF versus HFpEF joins a long list of characteristics that differ between these 2 major HF phenotypes. What clearly needs to be done in the heterogeneous HFpEF population is to identify phenotypic subtypes that exhibit pathophysiologic features that can be approached with precision therapy, potentially including a subgroup that would favorably respond to heart rate lowering.
↵∗ Editorials published in JACC: Heart Failure reflect the views of the authors and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology.
Dr. Bristow was the chairman of the TOPCAT (Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist) study Data and Safety Monitoring Board; and is the administrative principal investigator. Dr. Altman is the clinical principal investigator of the American Heart Association-sponsored PROBE-IT (Pulse Reduction On Beta-blocker and Ivabradine Therapy; NCT02973594) study where blinded study medication is being supplied by Amgen.
- 2017 American College of Cardiology Foundation
- Nagatsu M.,
- Spinale F.G.,
- Koide M.,
- et al.
- Kotecha D.,
- Flather M.D.,
- Altman D.G.,
- et al.
- O'Neal W.T.,
- Sandesara P.B.,
- Samman-Tahhan A.,
- Kelli H.M.,
- Hammadah M.,
- Soliman E.Z.
- ↵Vazir A, Claggett B, Pitt B, et al. Prognostic importance of temporal changes in resting heart rate in heart failure and preserved ejection fraction from the TOPCAT study. J Am Coll Cardiol HF. In press.