Esmolol infusion versus propranolol infusion: effects on heart rate and blood pressure in healthy volunteers

We used cycle ergometry exercise and infusions of isoproterenol and epinephrine to test the heart rate-lowering effect of esmolol compared with propranolol and saline in healthy humans. Collectively, our data indicate that infusion of ~160 mg of esmolol (range 110-200 mg in the 5 min before exercise) acutely and selectively blocks β₁-adrenergic receptors. These infusion parameters can be used in future experiments to evaluate β₁- vs. β₂-receptor control of the circulation in humans.

Abstract

Despite its widespread clinical use, the β₁-adrenergic receptor antagonist esmolol hydrochloride is not commonly used in human physiology research, and the effective dose of esmolol (compared with the nonselective β-blocker propranolol) is unclear. In four separate studies we used cycle ergometry exercise and infusions of isoproterenol and epinephrine to test the heart rate (HR)-lowering effect of esmolol compared with propranolol and saline in healthy humans. In cohort 1, both esmolol (ΔHR 57 ± 6 beats/min) and propranolol (ΔHR 56 ± 7 beats/min) attenuated exercise tachycardia compared with saline (ΔHR 88 ± 17 beats/min). In cohort 2, we found that the HR response to exercise was similar at 5 min (ΔHR 57 ± 9 beats/min) and 60 min (ΔHR 55 ± 9 beats/min) after initiation of the esmolol maintenance infusion. In cohort 3, we confirmed that the HR-lowering effect of esmolol disappeared 45 min after termination of the maintenance infusion. In cohort 4, changes in femoral blood flow and hematological parameters in response to epinephrine infusion were not different between esmolol and saline infusion, indicating that our esmolol infusion paradigm does not block β₂-receptors. Collectively, our data indicate that infusion of ~160 mg of esmolol (range 110-200 mg in the 5 min before exercise) acutely and selectively blocks β₁-receptors in healthy humans. Additionally, β₁-receptors remain blocked 60 min later if a maintenance infusion of ~0.2 mg·kg total body mass⁻¹·min−¹ continues. The current data lay the foundation for future studies to evaluate β₁- vs. β₂-receptor control of the circulation in humans.

NEW & NOTEWORTHY We used cycle ergometry exercise and infusions of isoproterenol and epinephrine to test the heart rate-lowering effect of esmolol compared with propranolol and saline in healthy humans. Collectively, our data indicate that infusion of ~160 mg of esmolol (range 110-200 mg in the 5 min before exercise) acutely and selectively blocks β1-adrenergic receptors. These infusion parameters can be used in future experiments to evaluate β1- vs. β2-receptor control of the circulation in humans.

the nonselective β-adrenergic receptor antagonist propranolol can be administered by intravenous infusion to evaluate neural control of the circulation in humans (23, 28, 37–39, 47, 59). However, because propranolol is nonselective, these prior studies cannot differentiate the effects of β₁- vs. β₂-receptors on the observed physiological responses. To determine β₁- vs. β₂-adrenergic effects, we first need to identify a cardioselective β₁-blocker that attenuates physiological and pharmacological increases in heart rate (HR) and contractility as effectively as propranolol [thereby isolating the β₂-receptor vascular component; the reduction in exercise HR parallels β₁-receptor occupancy (56)]. Ideally, this β₁-blocker would be commercially available for intravenous use, with good safety profile, have a large effect (pharmacodynamics), and have rapid onset and short context-sensitive half time (pharmacokinetics). From a clinical perspective, atenolol and metoprolol are thought to be relatively cardioselective β₁-blockers that might help settle this debate. However, at high doses, atenolol and metoprolol also block β₂-receptors (60, 63). Moreover, the half-life for both of these drugs is 4-6 h. Currently, the best β₁-blocker for use in human physiology research in unknown.

Esmolol hydrochloride (Brevibloc, Baxter Healthcare) is commonly used in clinical practice as an ultra-short-acting β₁-blocker (4, 19, 20, 30, 51). With an elimination half-life of <10 min due to metabolism by red blood cell esterases and low lipid solubility, it is often administered as a bolus injection before transient stimulation such as laryngoscopy, cranial pinning, and electroconvulsive therapy (5, 58). Indeed, comparative effectiveness studies in anesthetized patients have demonstrated that bolus injections of 20–200 mg of esmolol attenuate HR responses to laryngoscopy and endotracheal intubation (13, 14, 29, 33, 41, 43). Esmolol can also be administered as a continuous infusion to treat intraoperative tachycardia due to airway or surgical stimulation (10, 16, 18, 31, 61), as part of the hemodynamic management during resection of pheochromocytomas (40, 62), to control HR during acute coronary syndromes (3) and thyrotoxicosis (8, 12), and to aid in the diagnosis of supraventricular tachycardia (34). Studies in healthy volunteers conducted in the 1980s and 1990s demonstrated that esmolol infusion is effective (vs. no treatment) and comparable to propranolol at lowering exercise HR (11, 27, 42, 54). However, these prior studies have several limitations, including poor reporting of infusion parameters, no measurement of arterial blood pressure (BP), unclear onset and offset of β-blockade, and uncertainty about β₂-receptor involvement in the observed responses. To summarize, the dose and infusion parameters used in prior publications are widely variable (1, 9, 13, 24, 48, 49, 55) and may not be adequate to fully antagonize the β₁-adrenergic receptor. Because esmolol is short-acting, it would be an ideal β-blocker to use in human physiology research in which participants are commonly discharged home at the completion of the study. However, several unanswered questions remain.

The purpose of this study was to determine a safe and effective esmolol dosing regimen for use in human physiology research. We conducted a series of studies in healthy humans to test four related hypotheses. 1) Esmolol infusion elicits a reduction in exercise HR similar to that elicited by propranolol infusion. 2) The reduction in exercise HR is apparent at both 5 and 60 min after initiation of esmolol maintenance infusion. 3) The HR-lowering effect of esmolol disappears 45 min after termination of the maintenance infusion. 4) Changes in femoral blood flow and hematological parameters in response to epinephrine infusion are not different between esmolol and saline (control) infusion.

METHODS

Design and subjects.

These laboratory experiments used a repeated-measures, crossover design and the infusion (esmolol, propranolol, and normal saline) served as the independent variable. Physiological parameters were measured continuously during resting baseline, stressors (i.e., exercise or systemic infusions of adrenergic agonists), and recovery. All studies were conducted in the Clinical Research Center at Penn State University College of Medicine (temperature 20–22°C), and subjects fasted for ≥4 h before the experiments. Subjects also avoided caffeine, alcohol, and exercise for 24 h before the studies.

All study protocols were approved in advance by the Institutional Review Board of Penn State University College of Medicine and conformed to the Declaration of Helsinki. All participants voluntarily provided written and informed consent. Because participants served as their own control, we did not purposely recruit individuals of a specific age or sex, but we ensured that all were healthy and not taking any medications. Menstrual status was not documented. Demographic information is presented in Table 1. Both male and female participants were invited to participate in as many protocols as they were willing and able to perform. In total, two subjects participated in cohorts 1, 2, and 3, and two other subjects participated in cohorts 2 and 3.

Table 1.

Demographics for each cohort

	Cohort 1	Cohort 2	Cohort 3	Cohort 4
Sample, M/F	5/0	4/1	6/2	3/0
Age, yr	26 (25–29)	26 (25–62)	26 (23–67)	24 (22–26)
Height, m	1.80 (1.60–1.93)	1.75 (1.60–1.80)	1.79 (1.55–1.83)	1.83 (1.79–1.93)
Weight, kg	80.3 (55.6–97.6)	70.4 (55.6–80.2)	75.2 (58.1–101.9)	79.6 (64.2–92.3)
Body mass index	24.7 (21.7–29.1)	22.9 (21.7–24.7)	24.0 (22.5–30.4)	23.8 (20.0–24.8)
Body fat, %	16.4 (10.8–26.6)	16.8 (10.8–28.2)	19.7 (10.5–28.2)	10.5 (10.2–12.2)
Fat-free mass, kg	71.6 (40.8–80.9)	58.9 (40.8–71.6)	61.4 (45.0–74.4)	71.3 (57.7–81.0)
Esmolol loading, mg	36 (20–40)	88 (61–107)	92 (68–112)	107 (86–122)
Esmolol infused at 5 min, mg	107 (61–121)	162 (112–197)	169 (124–205)	196 (158–222)
Esmolol infused at 60 min, mg	895 (510–1,011)	972 (673–1,181)	1,012 (742–1,228)	1,176 (951–1,337)

Data are shown as median (minimum–maximum).

Physiological measurements.

A three-lead ECG (Cardiocap/5, GE Healthcare) and a finger BP cuff (Finometer, Finapres Medical Systems) were placed. Depending on the study, either one or two intravenous catheters were placed in the antecubital space. Prior to each stressor, three resting BPs were obtained by automated oscillometry of the right brachial artery (vital signs monitor SureSigns VS3, Philips) after 15 min of quiet rest and used to verify the Finometer values as previously described (38). All beat-by-beat variables were collected at 200 Hz by a data acquisition system (PowerLab, ADInstruments, New Castle, Australia). In some studies (cohort 3), arm BP was also measured by an automated ausculatory cuff (SunTech Tango) once per minute during exercise.

Screening visit.

Prior to experimental visits, the subjects attended a screening visit, during which they underwent a standard history and physical examination, resting echocardiogram, dual-energy X-ray absorptiometry (DXA) scan, fasting blood panels (lipids and comprehensive metabolic panel), and, finally, a maximal treadmill exercise test with respiratory gas measurement (ParvoMedics) and 12-lead ECG monitoring. Findings from all these tests were interpreted by a cardiologist before the subjects were enrolled in the study. Exclusion criteria were as follows: pregnant or nursing women, individuals with resting HR <45 beats/min, and history of cardiovascular, pulmonary, renal, or endocrine disease. We obtained a DXA scan so that we could administer medications based on fat-free mass (FFM), instead of total body mass, as FFM correlates with total blood volume (26) and esmolol is metabolized by red blood cell esterases (58).

Cohort 1.

Five young healthy men (Table 1) participated in cohort 1 in April–November 2014. This was a single-blind study, and trials were conducted in random order. Each treatment (esmolol, propranolol, and saline) occurred on a separate day and was separated by ≥1 wk. The experimental timeline for cohort 1 is depicted in Fig. 1. Resting HR and BP were determined in the supine posture, and a blood sample was obtained for the measurement of glucose, potassium, and lymphocytes (Hershey Medical Center Clinical Laboratories). Based on the literature and clinical guidelines (11, 27, 42, 54, 58), we chose to administer esmolol hydrochloride (2,500 mg in 250 ml) intravenously as 0.5 mg·kg FFM⁻¹·min⁻¹ for 1 min and then 0.2 mg·kg FFM⁻¹·min⁻¹ for the remainder of the study. Propranolol hydrochloride (50 mg in 245 ml of normal saline) was administered intravenously as a loading dose of 0.4 mg/kg FFM in 15 min followed by 1.3 µg·kg FFM⁻¹·min⁻¹ for the remainder of the study. The saline trial followed the same infusion parameters as the esmolol trial to account for time and volume effects. These drugs were always administered into the left arm by an infusion pump (Alaris PC model 8015, CareFusion).

Fig. 1.

Experimental timeline for cohort 1. Arrows denote when blood samples were obtained.

Isometric handgrip exercise at 40% maximal voluntary contraction was performed with the right hand until fatigue. All subjects were right-hand-dominant, and BP was measured on the left arm. Fatiguing isometric handgrip exercise raises sympathetic tone, HR, and BP (32, 35, 36). To provide visual feedback to the subject during handgrip, an analog meter display was used to interface a Lafayette dynamometer to a custom device. Muscular fatigue was defined as the inability to maintain the target workload despite verbal encouragement. After 15 min of rest, subjects pedaled on a recumbent bicycle (Nautilus NR 2000) at 60 rpm for 6 min. The workloads necessary to increase HR >60 beats/min above baseline were determined during the screening visit. The workload was progressive for all subjects (i.e., increasing every 1–3 min based on HR). Each subject required a different workload, and the workloads were kept constant for each treatment (saline, esmolol, and propranolol). HR and BP were measured on a beat-by-beat basis. After 15 min of rest, another blood sample was obtained from the left arm; then an 18-min isoproterenol hydrochloride (0.2 mg in 700 ml of saline) infusion was administered in the right arm as previously described (53). Isoproterenol is a nonselective β-agonist that has been used previously to challenge β-blockade with esmolol and/or propranolol (37, 38, 42). The first stage was 9 ng·kg FFM⁻¹·min⁻¹ for 3 min, and the dose was increased by 3 ng every 3 min until a level of 24 ng·kg FFM⁻¹·min⁻¹. The doses of isoproterenol are similar to those used in previous studies (44–46). HR and BP were measured on a beat-by-beat basis. A final blood sample was obtained within 2 min of the end of isoproterenol infusion.

Cohort 2.

Based on the data obtained from cohort 1, we increased the dose of esmolol and enrolled cohort 2 in May–August 2015. Five healthy subjects (Table 1) volunteered to participate. The dose of esmolol was 0.5 mg·kg FFM⁻¹·min⁻¹ for 3 min followed by 0.25 mg·kg FFM⁻¹·min⁻¹ for the remainder of the study. This dose of esmolol is nearly three times higher than that used in cohort 1 but is still within the clinical guidelines and the range used in prior publications (11, 27, 42, 54, 58). Similar to cohort 1, exercise HR and BP responses under esmolol blockade were compared with exercise HR and BP under propranolol blockade (same dose and timing as cohort 1). Additionally, we also tested the hypothesis that the reduction in exercise HR (compared with no β-blockade) is apparent at both 5 and 60 min after initiation of esmolol maintenance infusion. Figure 2 displays the procedures used in cohort 2. Subjects pedaled on a recumbent bicycle (Nautilus NR 2000) at 60 rpm for 6 min, and the workload was increased each minute. Consistent with prior studies, the maintenance infusion of both esmolol and propranolol ran for 5 min before the onset of exercise (42). Each exercise bout was separated by 45–50 min of rest (54). Because the half-life of esmolol is 9 min, we chose to wait five half-lives (45 min) between exercise bouts to confirm that our chosen maintenance dose effectively attenuated exercise HR. In practice, however, the time between the end of the second biking trial and the end of the third biking trial was ~60 min for each participant (due to baseline measurements and the length of the exercise bouts). Therefore, we refer to the exercise bouts as “esmolol-5” and “esmolol-60.”

Fig. 2.

Experimental timelines for cohorts 2–4. Arrows denote when blood samples were obtained.

Cohort 3.

Eight healthy subjects (Table 1) participated in additional exercise studies conducted in December 2015-May 2016 (cohort 3). These studies were single-blind and randomized. In contrast to cohorts 1 and 2, which used progressively increasing workloads, we used steady-state cycle ergometry in cohort 3. The 7-min exercise bouts began with a 30-s passive warm-up (to overcome the inertia of the machine) followed by 3 min of low-intensity pedaling at 60 rpm (target HR 110 beats/min) and, finally, 3.5 min of moderate-intensity pedaling at 60 rpm (target HR 150 beats/min). Exercise was performed before β-blockade, during β-blockade, and 45 min after termination of β-blockade (Fig. 2). We used the same doses of esmolol and propranolol used in cohort 2 and measured HR and BP as described above. For cohort 3, arm BP was also measured by an automated ausculatory cuff (SunTech Tango) once per minute during exercise. Because this cuff uses the ausculatory method, we reasoned that systolic BP and diastolic BP would be more accurate than values observed with the Finometer. We tested the hypothesis that the HR-lowering effect of esmolol disappears 45 min after termination of the maintenance infusion.

Cohort 4.

Three young healthy men (Table 1) participated in additional studies in October 2015–May 2016. In a single-blind and randomized manner, esmolol (same parameters as cohorts 2 and 3) and saline were infused into the left arm on separate days (Fig. 2). In the right arm, epinephrine (1 mg in 500 ml of saline) was infused over 32 min as previously described (53): 25 ng·kg FFM⁻¹·min⁻¹ in stage 1, 50 ng·kg FFM⁻¹·min⁻¹ in stage 2, 100 ng·kg FFM⁻¹·min⁻¹ in stage 3, and 200 ng·kg FFM⁻¹·min⁻¹ in stage 4. In addition to measuring HR and BP continuously, we also measured femoral blood flow velocity with Doppler ultrasound (Philips iE33); velocity waveforms were synchronized to a data acquisition system (PowerLab, ADInstruments) and analyzed offline (22). Femoral artery diameters were obtained each minute during the infusion, and femoral blood flow was calculated by multiplying the cross-sectional area (πr²) of the vessel by mean femoral velocity and by 60 to express flow in units of ml/min. We tested the hypothesis that changes in femoral blood flow and hematological parameters in response to epinephrine infusion are not different between esmolol and saline (control) infusion. We attempted to perform identical experiments under propranolol blockade but discontinued data collection because epinephrine infusion at 25 ng·kg FFM⁻¹·min⁻¹ caused a large rise in BP and HR values of 30–40 beats/min (i.e., below our safety threshold). Numerous published studies have shown that infusion of epinephrine during propranolol blockade causes unopposed α-adrenergic vasoconstriction, heightened BP, and reduced HR (6, 21, 25, 52, 57).

Data collection and statistical analysis.

All variables were monitored continuously, and data were analyzed offline at specific time points. Repeated-measures analysis of variance was used to determine the effect of treatment (saline, esmolol, and propranolol) on physiological variables over time. Planned comparisons were made using paired t-test at select time points (e.g., baseline and end exercise). We also calculated absolute and percent changes (Δ) for comparison with the literature. Specifically, we calculated the percent reduction in HR due to β-blockade as follows: (rise in HR without β-blocker – rise in HR with β-blocker)/rise in HR without β-blocker (42). To determine effect sizes for the primary outcomes, Cohen’s d values were calculated as the t-statistic/square root of the sample size. Cohen’s d values >0.8 were considered to be “strong” effect sizes. Perceptual values were compared between treatments with Wilcoxon's nonparametric test. Values are means ± SD unless otherwise stated. P < 0.05 was considered statistically significant.

RESULTS

Cohort 1.

Handgrip duration was similar during the saline (145 ± 17 s), esmolol (138 ± 17 s), and propranolol (145 ± 15 s) trials. During the saline trial, HR increased from 54 ± 7 to 75 ± 11 beats/min; during esmolol, the change in HR was attenuated (from 56 ± 9 to 67 ± 4 beats/min, P = 0.040 vs. saline, Cohen’s d = 1.06). During the propranolol trial, HR increased from 52 ± 4 to 62 ± 4 beats/min, which was significantly less than during the saline trial (P = 0.022, Cohen’s d = 1.32) but similar to the esmolol trial (P = 0.955). The handgrip-mediated change in mean arterial pressure (MAP) was similar during saline (34 ± 9 mmHg), esmolol (31 ± 6 mmHg), and propranolol (34 ± 8 mmHg) trials. Taken together, the handgrip data suggest that esmolol and propranolol at the chosen infusion rates have similar effects on exercise HR but the absolute end-exercise HR values were rather low (i.e., far below 100 beats/min) and, therefore, indicate submaximal activation of the sympathetic nervous system.

As shown in Fig. 3, the HR response to bicycle exercise was attenuated by both esmolol and propranolol compared with saline (P = 0.003, Cohen’s d = 2.34). The percent reduction in exercise HR was 33 ± 10% with esmolol and 34 ± 10% with propranolol, indicating that reflex tachycardia is similarly affected by these two drugs. However, because resting HR was lower with propranolol than esmolol (P = 0.021), the end-exercise HR was also lower with propranolol than esmolol (P = 0.038). The change in Finometer-derived MAP at the end of bicycle exercise was higher during the propranolol (35 ± 10 mmHg) than the esmolol (18 ± 10 mmHg) trial (P = 0.044). The change in MAP during the saline trial was 29 ± 22 mmHg, which was not different from the esmolol (P = 0.149) or the propranolol (P = 0.450) trial. The median rating of perceived exertion obtained at the end of exercise was 16 (range 14–20) during saline infusion, 17 (range 14–19) during esmolol infusion, and 19 (range 16–20) during propranolol infusion (Wilcoxon's Z = −2.0, P = 0.048 vs. esmolol).

Fig. 3.

Absolute heart rate (top) and changes (Δ) in heart rate (bottom) in response to 6 min of recumbent bicycle exercise in cohort 1 (i.e., low-dose esmolol). Error bars have been omitted for clarity. See results for statistical comparisons between treatments.

The HR responses to isoproterenol are depicted in Fig. 4. As expected, there was no change in HR under propranolol. During stage 6 of the isoproterenol infusion, absolute HR was not different between esmolol and saline trials (P = 0.370). The change in HR was higher with esmolol than propranolol (Fig. 3, bottom; P < 0.001, Cohen’s d = 2.85). During the saline trial, Finometer-derived MAP was reduced by isoproterenol (−9 ± 5 mmHg, P < 0.001 vs. baseline), but this depressor response was not observed during the esmolol (0 ± 4 mmHg change from baseline) or the propranolol (2 ± 3 mmHg change from baseline) trial. In response to isoproterenol infusion during the saline trial, blood glucose (5% increase), potassium (3% decrease), and lymphocytes (24% increase) changed very little, so we decided to not statistically analyze the esmolol and propranolol blood samples.

Fig. 4.

Heart rate responses to increasing doses of isoproterenol (top) and changes (Δ) in heart rate from baseline to the last stage of the infusion (bottom). Error bars have been omitted for clarity. See results for statistical comparisons between treatments.

Cohort 2.

The absolute and relative HR responses to bicycle exercise are displayed in Fig. 5. In the control unblocked condition (pre-esmolol and pre-propranolol, which occurred on separate days), HR responses were very similar (Fig. 5, bottom). The percent reduction in exercise HR (relative to pre-esmolol) was 36 ± 6% during the first esmolol trial, 38 ± 7% during the second esmolol trial, and 34 ± 6% with propranolol, indicating that reflex tachycardia was similar regardless of treatment. It should be noted that propranolol lowered resting HR compared with pre-esmolol (P = 0.011) and esmolol (P = 0.001). Therefore, it follows that the end-exercise HR was lower with propranolol than in the esmolol (P = 0.004) or pre-esmolol (P = 0.002) trial. Compared with the pre-esmolol trial (41 ± 22 mmHg), the changes in MAP (Finometer-derived) were not different between the first esmolol trial (26 ± 17 mmHg, P = 0.062), second esmolol trial (32 ± 16 mmHg, P = 0.265), and propranolol trial (35 ± 11 mmHg, P = 0.468). Together these exercise data indicate that 1) the degree of β-blockade in our subjects was similar at 5 and 60 min after initiation of the esmolol maintenance infusion and 2) the degree of β-blockade was similar between esmolol and propranolol trials.

Fig. 5.

Absolute heart rate (top) and changes (Δ) in heart rate (bottom) in response to 6 min of recumbent bicycle exercise in cohort 2 (i.e., high-dose esmolol). Error bars have been omitted for clarity. Because heart rate responses to pre-esmolol and pre-propranolol were nearly identical, average responses are presented graphically as a single “unblocked” line. See results for statistical comparisons between treatments. Esmolol-5 and esmolol-60, esmolol infusion at 5 and 60 min.

The median rating of perceived exertion obtained at the end of exercise was 16 (range 14–20) during the pre-esmolol trial, 18 (range 15–19) during the first esmolol trial (Wilcoxon's Z = −1.41, P = 0.157 vs. saline), 18 (range 15–19) during the second esmolol trial (Wilcoxon's Z = −1.30, P = 0.194 vs. saline), and 19.5 (range 19–20) during the propranolol trial (Wilcoxon's Z = −1.63, P = 0.051 vs. saline and Wilcoxon's Z = −1.84, P = 0.033 vs. esmolol). The participants did not experience adverse events such as hypotension, lightheadedness, or nausea during any of the trials.

Cohort 3.

Hemodynamics at rest and in response to steady-state exercise are displayed in Table 2. Esmolol infusion (n = 8) attenuated the increase in exercise HR compared with pre-esmolol (P < 0.001, Cohen’s d = 4.41). The end-exercise HR was 138 ± 14 beats/min during pre-esmolol, 119 ± 12 beats/min during esmolol infusion, and 138 ± 15 beats/min 45 min after termination of esmolol infusion. Stated another way, the percent reduction in exercise HR (relative to the preblockade exercise bout) was 29 ± 7% during esmolol infusion and 15 ± 7% when esmolol infusion was turned off (P = 0.004, Cohen’s d = 1.48). Esmolol also attenuated the increase in systolic BP due to exercise (P < 0.001 vs. pre-esmolol, Cohen’s d = 3.07), and this effect disappeared 45 min after the esmolol infusion was turned off (P = 0.136 vs. pre-esmolol).

Table 2.

Resting and exercise hemodynamics for cohort 3

	Resting HR, beats/min	Resting Systolic BP, mmHg	Resting Diastolic BP, mmHg
Pre-esmolol	69 ± 8	104 ± 8	68 ± 6
Esmolol	70 ± 7†	105 ± 8	74 ± 4†
45 min post-esmolol	80 ± 8*	104 ± 7	68 ± 3
Pre-propranolol	65 ± 7	113 ± 10	68 ± 7
Propranolol	58 ± 5	101 ± 8*	69 ± 9
45 min post-propranolol	60 ± 5	110 ± 13	70 ± 6

	ΔHR, beats/min	ΔSystolic BP, mmHg	ΔDiastolic BP, mmHg
Pre-esmolol	68 ± 11	76 ± 15	8 ± 7
Esmolol	49 ± 12*†	40 ± 10*†	−1 ± 8*†
45 min post-esmolol	58 ± 10*	68 ± 10	9 ± 8
Pre-propranolol	70 ± 17†	65 ± 17†	0 ± 7
Propranolol	52 ± 16*	36 ± 17*	9 ± 10*
45 min post-propranolol	50 ± 13	34 ± 24	8 ± 14

Values are means ± SD. Blood pressure (BP) was measured by an automated ausculatory BP device (SunTech Tango). Changes (Δ) from baseline in heart rate (HR) and BP were calculated using values obtained within the last 30 s of exercise. Sample size was 8 for the esmolol trial and 4 for the propranolol trial; therefore, comparisons are not made between treatments.

^*P < 0.05 vs. predrug on the same day;

^†P < 0.05 vs. 45 min postdrug on the same day.

On a separate day, a subset of individuals (n = 4) completed the propranolol trial (Table 2). The end-exercise HR was 136 ± 20 beats/min during pre-propranolol, 110 ± 14 beats/min during propranolol infusion, and 110 ± 14 beats/min at 45 min after the propranolol infusion was turned off. Together these exercise data confirm that the context-sensitive half time of propranolol is several magnitudes higher than that of esmolol. Additionally, the participants did not experience adverse events such as hypotension, lightheadedness, or nausea during any of the trials.

Cohort 4.

In response to systemic epinephrine infusion, the change in HR was greater during the saline (17 ± 1 beats/min) than the esmolol (1 ± 3 beats/min) trial (P = 0.031, Cohen’s d = 3.24). The increases in femoral blood flow were not attenuated by esmolol (138 ± 63 ml/min) compared with the saline control trial (96 ± 95 ml/min, P = 0.595). The percent decrease in potassium was not different between esmolol (−19 ± 7%) and saline (−26 ± 9%, P = 0.188). In a similar way, the percent increase in blood glucose was not different between esmolol (75 ± 24%) and saline (78 ± 30%, P = 0.535). The percent increase in lymphocyte count was also not different between esmolol (133 ± 39%) and saline (130 ± 18%, P = 0.934). Together the data from cohort 4 indicate that our dose of esmolol does not block β₂-adrenergic receptors, because epinephrine infusion elicited similar femoral artery blood flow, similar hypokalemia, similar hyperglycemia, and similar lymphocytosis between treatments.

DISCUSSION

The purpose of this study was to determine a safe and effective esmolol dosing regimen for use in human physiology research. Our data indicate that a loading dose of 0.5 mg·kg FFM⁻¹·min⁻¹ for 3 min followed by a maintenance infusion of 0.25 mg·kg FFM⁻¹·min⁻¹ for the remainder of the study is safe, effective, and selective for β₁-blockade with rapid onset and offset. These data confirm and extend on prior investigations and provide a framework for future human physiology studies.

Adrenergic receptors play an important role in blood flow regulation in both health and disease. The myocardial β₁-receptors are particularly important during exercise, because HR must increase so that cardiac output can increase perfusion to working skeletal muscle. β₁-Receptors are also involved in lipolysis (2) and renin release (17). However, only a handful of prior investigators have used esmolol infusion to study blood flow regulation in humans; these studies either used single bolus injections (9, 15) or did not state how long the maintenance infusion continued (27, 50) or did not compare esmolol with propranolol (24, 54, 55). Indeed, there is no established “cardioselective” dose of esmolol in the published literature. The current data indicate that not only does esmolol attenuate the rise in exercise tachycardia as effectively as propranolol, but it also does not provoke hypotension or nausea as suggested by prior investigators (11, 42). Additionally, we observed that the β₁-blocking effect of esmolol disappears ~45 min after termination of the maintenance infusion (Table 2), whereas esmolol continues to be effective (vs. propranolol) if the maintenance infusion continues (Fig. 5). In cohort 4, we determined that our esmolol infusion does not block β₂-receptors by showing that the changes in femoral blood flow and hematological parameters in response to epinephrine infusion are similar between esmolol and saline (control) infusion.

Our data from cohort 1 and the published literature indicate that exercise and isoproterenol infusion have different effects on β₁-receptors. Similar to our data (Fig. 4), Reilly et al. (42) found that even the highest doses of esmolol (0.3-0.75 mg·kg¹·min⁻¹ maintenance infusion) could not block isoproterenol-induced tachycardia to the same degree as propranolol. If these isoproterenol data are viewed in isolation, it could be inferred that esmolol is not as efficacious as propranolol at antagonizing β₁-receptors. However, we believe that this is not the case. When isoproterenol is infused systemically, it stimulates both β₁- and β₂-receptors in the heart (raising HR and contractility) and peripheral blood vessels (causing vasodilation). Under β₁-blockade with esmolol, the observed HR responses to isoproterenol are likely due to baroreflex-mediated vagal withdrawal, but stimulation of unblocked cardiac β₂-receptors can also raise HR (7). Some authors have suggested that exercise HR responses are more indicative of β₁-stimulation by norepinephrine (which is more selective for β₁-receptors), whereas HR responses to isoproterenol reflect both β₁- and β₂-receptor activation (42). To summarize, our data support the previous literature that HR responses to isoproterenol are more effectively blocked by nonselective β-blockers than β₁-blockers, even at doses that produce equal suppression of exercise HR.

Limitations.

The current study was designed to evaluate HR and BP responses to exercise and pharmacological infusions but is underpowered to evaluate HR at rest and ratings of perceived exertion (which appear to be different between esmolol and propranolol). Because most patients undergoing cardiac rehabilitation are treated with β-blockers (i.e., metoprolol and atenolol, which are β₁-selective, or carvedilol, which is nonselective), our finding that ratings of perceived exertion are lower with esmolol than with propranolol (cohort 2) may have clinical relevance but require appropriately designed and adequately powered future investigations. Because this was a crossover study evaluating effects of different β-blockers, we enrolled both men and women of various age categories. This approach prevents us from determining age and sex differences. Another limitation of the current study is that we do not have a measurement of stroke volume, contractility, or cardiac filling pressures. Additionally, the sample size for cohort 4 is low, but the main finding (i.e., lack of β₂-blocking effect) is unlikely to change with enrollment of more subjects. Finally, the absolute HR values at end exercise were higher with esmolol than with propranolol (Fig. 5), which may indicate that β₂-receptors contribute to exercise tachycardia, but new experimental approaches are needed to confirm or refute this claim.

Conclusions.

Esmolol is a short-acting β₁-blocker that can be infused intravenously to attenuate HR responses to sympathetic stress. It has several clinical uses but has only been used sparingly in human physiology research. We conducted four separate studies with esmolol and found that a loading dose of 0.5 mg·kg FFM⁻¹·min⁻¹ for 3 min followed by a maintenance infusion of 0.25 mg·kg FFM⁻¹·min⁻¹ for the remainder of the study is safe, effective, and selective for β₁-blockade with rapid onset and offset. The current data lay the foundation for future studies to evaluate β₁- vs. β₂-receptor control of the circulation in humans.

GRANTS

This project was supported by a grant from the Association of Faculty and Friends of the Penn State Milton S. Hershey Medical Center (M. D. Muller). This project was also supported, in part, by National Center for Advancing Translational Sciences Grants UL1 TR-000127 and KL2 TR-000126 and also under a grant with the Pennsylvania Department of Health using Tobacco CURE funds (M. D. Muller).

DISCLAIMERS

The Pennsylvania Department of Health and the National Institutes of Health specifically disclaim responsibility for any analyses, interpretations, or conclusions.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

M.D.M., T.A.A., U.A.L., and P.M.M. conceived and designed research; M.D.M., T.A.A., A.F.V.P., D.N.P., J.C.L., A.J.R., and U.A.L. performed experiments; M.D.M., J.C.L., and A.J.R. analyzed data; M.D.M., T.A.A., A.F.V.P., D.N.P., A.S.B., J.C.L., S.R.M., A.J.R., U.A.L., and P.M.M. interpreted results of experiments; M.D.M. prepared figures; M.D.M., A.F.V.P., D.N.P., A.S.B., and A.J.R. drafted manuscript; M.D.M., T.A.A., A.F.V.P., D.N.P., A.S.B., J.C.L., S.R.M., A.J.R., U.A.L., and P.M.M. edited and revised manuscript; M.D.M., T.A.A., A.F.V.P., D.N.P., A.S.B., J.C.L., S.R.M., A.J.R., U.A.L., and P.M.M. approved final version of manuscript.