Differential effects of eplerenone versus amlodipine on muscle metaboreflex function in hypertensive humans

Abstract Numerous studies have demonstrated that sympathetic nervous system overactivation during exercise in hypertensive rodents and humans is due, in part, to an exaggerated reflex response known as the exercise pressor reflex. Our prior studies have implicated a key role of mineralocorticoid receptor activation in mediating an augmented exercise pressor reflex in spontaneously hypertensive rats, which is mitigated by blockade with eplerenone. However, the effect of eplerenone on exercise pressor reflex has not been assessed in human hypertension. Accordingly, the authors performed a randomized crossover study to compare the effects of eplerenone to another antihypertensive drug from a different class amlodipine on sympathetic nerve activity (SNA) in 14 patients with uncomplicated hypertension. The authors found that amlodipine unexpectedly augmented the increase in SNA during the second minute of isometric handgrip, which persisted into the post‐exercise circulatory arrest period (∆ SNA, from rest of 15 ± 2 vs. 9 ± 2 vs. 10 ± 2 bursts/min, amlodipine vs. baseline vs. eplerenone, respectively, p < .01), suggesting an exaggerated muscle metaboreflex function. Eplerenone did not alter sympathetic responses to exercise or post‐exercise circulatory arrest in the same hypertensive individuals. In conclusions, our studies provide the first direct evidence for a potentially unfavorable potentiation of muscle metaboreflex by amlodipine during isometric handgrip exercise in hypertensive patients whereas eplerenone has no significant effect. Our study may have clinical implications in terms of selection of antihypertensive agents that have the least detrimental effects on sympathetic neural responses to isometric exercise.


INTRODUCTION
A large body of evidence in rodents and humans have indicated that hypertension is associated with overactivation of the sympathetic nervous system during exercise, which is independent of resting sympathetic nerve activity (SNA). [1][2][3][4] Normally, exercise produces intensitydependent increases in heart rate (HR), cardiac output (CO), and blood pressure (BP), mediated at least in part by increases in SNA. A reflex arising in the contracting skeletal muscle, known as exercise pressor reflex (EPR), is known to play a major role in driving SNA during exercise. The EPR can be activated either by metaboreceptors, which are stimulated slowly by the metabolic byproducts of muscle contraction (ie, metaboreflex), or mechanoreceptors, which respond quickly to mechanical deformation of the muscle fibers (ie, mechanoreflex). 5 Studies in young spontaneously hypertensive rat (SHR), a standard rat model of primary hypertension before development of heart failure, demonstrated that both the metabo-and mechanoreflex is overactive. 4,6 Studies in hypertensive patients have also suggested augmented EPR, 7 particularly in muscle metaboreflex function. 8,9 While the precise mechanisms underlying this observation are unknown, an increasing number of studies have demonstrated a potentially important role for central mineralocorticoid receptors (MR) in mediating sympathetic nervous system overactivation even in the presence of normal circulating aldosterone levels. 10,11 Furthermore, our recent studies have implicated a key role for MR activation in mediating the augmented EPR observed in hypertensive rats 12 which was reversed by treatment with the MR antagonist eplerenone. 10 However, the role of MR in modulating EPR function in human hypertension has not been assessed.
Therefore, we conducted a randomized crossover study to assess the effects of eplerenone, a selective MR antagonist, on SNA and BP during static and dynamic exercise in patients with uncomplicated hypertension. To further determine if the potential effects of eplerenone are specific to its drug class, we likewise compared the effects of eplerenone to the commonly used antihypertensive drug amlodipine, a calcium channel blocker known to have neutral effects on resting SNA. 13

METHODS
All experimental protocols were approved by the Institutional Review Board of the University of Texas Southwestern Medical Center. Written, informed consent was obtained from all of the patients.

Patients
Fourteen of which were given in combination of hydrochlorothiazide and two in combination with amlodipine. One patient treated with amlodipine alone, one with metoprolol alone, one with enalapril alone, and one with losartan alone. None of patients were treated with MR antagonists before the study.

Measurement of sympathetic nerve activity by microneurography
Multiunit recordings of SNA were obtained with unipolar tungsten microelectrodes inserted into muscle fascicles of the peroneal nerve by microneurography. 12 Neural signals were amplified, filtered (bandwidth 700-2000 Hz), rectified, and integrated to obtain mean voltage neurograms. Recordings were considered acceptable based on welldefined criteria that discriminate muscle SNA from other neural signals including skin SNA and muscle spindle activity. 13 Muscle SNA was expressed as burst frequency (bursts/min) and total activity (burst frequency × mean burst amplitude).

Cardiac output
Cardiac output was measured at rest and during handgrip exercise by thoracic electrical bioimpedance (BioZ, CardioDynamics, San Diego, CA, USA) as previously described. 14 Stroke volume (SV) was derived from change in impedance/time measured during electrical systole.
Cardiac output was determined as the product of SV and HR.

Test protocols
All patients underwent measurement of BP, HR, CO, SV, total periph-

STATISTICAL ANALYSIS
Mixed-effects linear models were used to conduct the repeated measures analysis to assess differences between the baseline period, amlodipine, and eplerenone phases. Time points measured during static hand grip, arm cycling, or CPTs were included in the mixedmodels as an added repeated factor. The model covariance structure was evaluated and selected based on Akaike Information Criterion.
Contrasts from these models were used for pair-wise comparisons.
Treatment order was also assessed in the models and no effect of treatment order on any outcome variables was found (all interaction

RESULTS
Baseline characteristics and biochemical changes during treatment with amlodipine and eplerenone are shown in Table 1. The average daily dose of amlodipine and eplerenone used in the study was 5.5±0.7 mg and 119.6±16 mg, respectively.
Serum potassium (K) was significantly increased after eplerenone treatment when compared to amlodipine. Diastolic BP tended to be lower after eplerenone and amlodipine but the difference did not reach statistical significance (Table 1). There was no change in body weight, body mass index, serum Cr, serum Na, fasting plasma glucose, total cholesterol, HDL-cholesterol, or triglyceride levels during either drug treatment.
Microneurographic recording of SNA was unsuccessful in one participant during eplerenone phase; however, all other available data for this participant was included in the analyses. Both eplerenone and amlodipine had no effect on resting SNA, SV, CO, TPR, or MAP (Table 2).
Eplerenone induced a small but significant increase in resting HR while amlodipine had no significant effect (Table 2). At baseline without antihypertensive drug treatment, static handgrip induced a significant increase in SNA, HR, and MAP in hypertensive patients (Figures 1-3).
As expected, when circulation to the exercising arm was prevented by cuff inflation above systolic BP at handgrip end, HR fell towards resting values during minute 3 and 4 while muscle SNA, and MAP remained elevated due to muscle metaboreflex activation. 15 There were no significant changes in TPR or SV during static handgrip or PECA. Cardiac output tended to increase during the first minute of handgrip (minute 1) and PECA (minute 4, p < .05 vs. minute 0, Figure 3B), but the overall mixed model p value is not significant. Eplerenone had no effect on Data are mean value ± standard deviation. Abbreviations: MAP, mean arterial pressure; HR, heart rate; SNA, sympathetic nerve activity. a For comparison between baseline and two treatment phases.
hypertensive patients ( Figure 3A). The increase in TPR during static handgrip tended to be augmented with amlodipine (pairwise p value < .05 vs. no drug during the first minute ( Figure 2D (Table 2).

DISCUSSION
The major findings of the study are two-fold. First, the selective MR antagonist eplerenone had no significant effects on SNA and BP responses to static and dynamic arm exercise in hypertensive patients.
Second, a commonly used calcium channel blocker, amlodipine, unexpectedly potentiated the rise in SNA and HR during static handgrip and post-handgrip circulatory arrest (a metaboreflex isolating experimental procedure). This effect of amlodipine was not observed during dynamic arm cycling exercise, suggesting a selective metaboreflex sensitization during static exercise.
The mechanisms by which amlodipine induced sympathetic activation during handgrip exercise are unknown. Amlodipine is a long-acting dihydropyridine calcium channel blocker considered to be the one of the first-line drug treatments for hypertension. 17 We chose amlodipine as a control due to prior report of neutral effects on resting SNA in humans. 18 In our study, resting SNA was also unaffected by amlodip- peripheral resistance. 22,23 In the presence of hypertension, however, metaboreflex-induced increases in CO and SV are markedly attenuated and elevated TPR plays a larger role in maintaining BP during exercise. 24 Our study extended previous observations by demonstrating that amlodipine further exaggerated the rise in TPR during isometric exercise, which may be deleterious if sustained over a prolonged period in hypertensive individuals.
The mechanisms underlying the amlodipine-induced sensitization of the metaboreflex during static exercise remain unknown. Amlodipine may induce greater increases in muscle lactate or acidosis, which are major stimulators of group IV muscle afferents during static handgrip. However, this is unlikely as previous studies did not show an augmented increase in lactate production in isolated guinea pig hearts subjected to coronary ischemia in the presence of amlodipine. 25 Similarly, Gillies and coworkers demonstrated no significant impact of amlodipine on plasma lactate levels during leg cycling exercise in hypertensive patients. 26 Interestingly, in the same study, amlodipine was found to reduce resting BP without lowering BP during isometric exercise which is similar to the findings of our present study. 26 Amlodipine is an antagonist of the voltage-dependent L-type calcium channel (LTCC). At least four isoforms of LTCC have been identi- Our study demonstrated no significant effects of eplerenone in modulating sympathetic neural and pressor response to both isometric and dynamic exercise in hypertensive patients which is in contrast with the sympatho-inhibitory effect of this drug class demonstrated in our previous rodent studies. 10 There are several potential explanations underlying this discrepancy. First, the dose of eplerenone used in our study may be too low as it had no significant impact on resting BP though its effect on serum potassium was detected, suggesting its action on the distal nephron. Although studies suggest that eplerenone crosses the blood brain barrier in rodent models of hypertension and heart failure when given systematically, 10 less potent than spironolactone in lowering BP, though more selective to MR than spironolactone. 31 We have previously demonstrated that spironolactone mitigates EPR overactivity in SHR to the same degree as eplerenone. 9 Given this, we did not treat the hypertensive patients participating in this study with spironolactone. Unlike rodents, it is possible that spironolactone may have a greater effect in humans and should be examined in future studies. Second, the potential action of eplerenone on SNA may be evident only when the renin angiotensin system is activated. Our previous studies in hypertensive patients showed that spironolactone alone had no effects on the resting SNA while chlorthalidone, a thiazide diuretic proposed to be a first line diuretic for the treatment of hypertension, induced persistent sympathetic activation in the same hypertensive patients. 32 However, spironolactone exerts a sympatho-inhibitory action when given in combination with chlorthalidone in hypertensive patients. 33 Whether eplerenone exerts a similar influence on SNA at rest or during exercise when given in combination with thiazide diuretics (ie, during renin angiotensin system activation) remains to be evaluated. A meta-analysis of randomized controlled trials showed that BP was reduced by 3.5/2.5 mm Hg with moderate to high intensity endurance training with duration between 100 and 150 min per week. 37 In the same meta-analysis, the average BP reduction appears to be larger with isometric resistance training by 10.9/6.2 mm Hg, though the total number of studies were relatively smaller than the endurance training exercise. 38 Accordingly, a scientific statement from the American Heart Association (AHA) has adopted isometric handgrip and resistance exercise training in addition to endurance exercise training as acceptable modalities of exercise intervention to reduce BP. 35

CONCLUSIONS
We found that amlodipine, a first line antihypertensive agent, potentiated the rise in SNA and HR during static handgrip and post-handgrip circulatory arrest. This effect of amlodipine was not observed during dynamic arm cycling exercise, suggesting selective metaboreflex sen-sitization during static exercise. Contrary to preclinical studies, mineralocorticoid receptor antagonism did not have a significant effect on SNA during static or dynamic exercise in hypertensive humans. Further studies are needed to determine if the effect of amlodipine on metaboreflex represents a class effect and whether higher doses of eplerenone are needed to alter SNA during exercise in humans.