Comparison of Candoxatril and Atrial Natriuretic Factor in Healthy Men
Effects on Hemodynamics, Sympathetic Activity, Heart Rate Variability, and Endothelin
Abstract The purpose of these experiments was to compare the effects of endopeptidase inhibition with oral candoxatril on systemic and forearm hemodynamics and muscle sympathetic nerve activity with responses to a low-dose atrial natriuretic factor infusion. Eleven healthy men received at random on three separate days either intravenous saline, atrial natriuretic factor (1.6 pmol/kg per minute) plus saline, or oral candoxatril (200 mg) plus saline. Measurements were made at baseline and 30, 60, and 90 minutes after interventions. Atrial natriuretic factor lowered diastolic pressure (P<.01), central venous pressure (P<.001), forearm blood flow (P<.05), and forearm vascular compliance (P<.05) but had no effect on systolic pressure, heart rate or its variability, stroke volume, sympathetic nerve activity, plasma norepinephrine, or endothelin-1. Plasma epinephrine increased (P<.01). Candoxatril lowered central venous pressure (P<.001) and increased systolic pressure (from 116±6 to 120±7 mm Hg; P<.05), endothelin (from 4.6±1.1 to 6.8±3.2 pmol/L; P<.02), and epinephrine (P<.05), without affecting any other variables. Candoxatril and atrial natriuretic factor lowered central venous pressure in healthy men without causing a reflex increase in sympathetic nerve activity or norepinephrine, yet epinephrine rose. This suggests that both interventions may specifically inhibit sympathetic nerve traffic to muscle at physiological plasma atrial natriuretic factor concentrations. However, whereas the peptide lowered blood pressure, candoxatril increased systolic pressure. These contrasting hemodynamic responses may be related to differences in plasma atrial natriuretic peptide concentration and to altered endothelin metabolism by candoxatril.
- blood pressure
- cardiac output
- natriuretic peptide
- nervous system
- vascular resistance
We have previously documented a relative inhibitory effect of ANF on postganglionic MSNA of healthy men. This could not be explained by augmented baroreceptor afferent input but was consistent with either a central or ganglionic sympathetic inhibitory action.1 2 In those experiments ANF infusion increased its plasma concentration 20- to 30-fold above baseline values, that is, to levels well above the physiological range for healthy young men. The purpose of the present experiments was to assess the effects of increasing ANF concentration within the physiological range on cardiac and systemic hemodynamics and on direct and indirect indices of sympathetic nervous system activity. Responses to atriopeptidase inhibition with oral candoxatril3 4 5 6 7 8 9 were compared with responses to a low-dose ANF infusion, or to a saline infusion as a vehicle, or time control for these two interventions according to a random double-blind study design.
We studied 11 healthy young men, aged 21.4±1.3 years (mean±SD), each on 3 separate study days. Their mean weight was 72.2±7.9 kg; their mean height was 178.6±4.2 cm. A medical history, physical examination, and laboratory investigations excluded hypertension, concurrent illness, and medication use. These protocols were approved by the Human Subjects Review Committee of the University of Toronto, and informed written consent was obtained from all participants.
All experiments were performed at the same time of day. Subjects were studied supine. HR was derived from lead II of the electrocardiogram. Blood pressure was measured by an automatic cuff recorder.1 An intravenous catheter was placed in a left forearm vein for infusions. After local anesthesia a central venous catheter was introduced into an antecubital vein and advanced to an intrathoracic position. A pneumobelt was wrapped around the lower thorax.
Multiunit recordings of postganglionic MSNA were obtained from the peroneal nerve.1 CVP and respiratory excursions were monitored by Statham P23ID pressure transducers (Gould Inc) and recorded simultaneously onto paper, along with HR, the electrocardiogram, forearm volume, and the sympathetic mean voltage neurogram. Sympathetic bursts were identified by inspection of the mean voltage neurogram and expressed in terms of burst frequency (bursts per minute), burst incidence (bursts per 100 heartbeats), and units of integrated sympathetic nerve activity (the product of burst frequency and mean burst amplitude in millimeters). For comparison of these Units on different study days the average burst amplitude over each time period was referenced against the maximal amplitude observed on that day during an internal standard (such as apnea or Valsalva’s maneuver) to obtain a normalized integrated value for MSNA.1
Stroke volume (SV) and cardiac output (CO) were calculated with Doppler echocardiography as previously described.1 10 Systemic vascular resistance (SVR) was derived from CO, mean arterial pressure (MAP), and CVP as SVR (dyne · s · cm−5)=80×(MAP−CVP) (mm Hg)/CO (L/min).
FBF was estimated by venous occlusion plethysmography.1 FVR (expressed as resistance units) was calculated by dividing (MAP−CVP) by the mean of four to six measures of FBF (milliliters per minute per 100 milliliters of forearm volume). For estimation of FVC, pressure in the upper cuff was increased stepwise from 0 to 10, 20, 30, and 40 mm Hg, and each level of cuff pressure was maintained constant for at least 2 minutes until a stable forearm volume was attained. Changes in forearm venous volume from baseline were related to the applied venous transmural pressure and expressed as milliliters per 100 milliliters per millimeter of mercury.11 12
Spectral analysis of HRV followed the coarse graining method of Yamamoto and Hughson13 14 as detailed in a recent publication from our laboratory.15 HR data over 7 minutes during baseline and each intervention period were subjected to spectral analysis.
Plasma norepinephrine and epinephrine concentrations were determined by high-performance liquid chromatography with electrochemical detection.1 Plasma ANF concentration was determined after extraction (mean recovery, 77%) by radioimmunoassay.16 The interassay and intra-assay coefficients of variation of this method are 9% and 7%, respectively. The sensitivity of this assay is 0.4 pmol/L. The second messenger of ANF, cGMP, was measured by radioimmunoassay with a sensitivity of 0.5 nmol/L. Plasma ET-1 concentration was determined after extraction (recovery, 60% to 70%) by radioimmunoassay (Peninsula Laboratories). The sensitivity of this method is 1.25 pg/mL.
Each subject was studied three times at least 1 week apart. Subjects received at random either saline (0.76 mL/min) plus 2% albumin (as vehicle or time control for ANF), ANF (1.6 pmol/kg per minute, 50 μg/mL; BioChem ImmunoSystems Inc) reconstituted in saline plus 2% albumin, or saline plus oral candoxatril (200 mg; Pfizer Central Research) once baseline measurements were obtained. One of us (B.L.S.) was solely responsible for allocation, preparation of solutions, and administration of interventions and was alone with the subject when these were given. All other investigators were blinded to the allocation for each study day and to plasma ANF concentrations until all studies and analyses were completed.
The study was divided into four time periods: a 30-minute period of baseline data collection, then three periods spanning the 30-, 60-, and 90-minute marks after the start of each intervention. Mean values for blood pressure, HR, CVP, and MSNA were the average of 3 to 5 minutes straddling each of these periods.
Data are expressed as mean±SD. Changes in hemodynamics and sympathoneural and humoral variables from baseline values on each study day were assessed by one-way repeated-measures ANOVA (systat). Paired t tests were used for assessment of the effects of each intervention on ET-1 concentration. Two-factor repeated-measures ANOVA (drug, time) was used for assessment of the effect of active interventions over time against saline. We applied the Friedman test to the frequency domain data since these were not distributed normally. A value of P<.05 was required for statistical significance.
Ten of the 11 subjects completed all 3 study days. The 11th subject did not attend the third study, at which time he would have received candoxatril.
Saline alone had no significant effect on systemic blood pressure or CVP but increased HR, stroke volume, and cardiac output (Table 1⇓, Figs 1⇓ and 2⇓). FBF, FVR, and FVC were also stable during the saline infusion (Table 1⇓). Nerve traffic could not be recorded in 1 of these subjects. Stable mean voltage neurograms were obtained throughout the entire 120-minute protocol in 9 of the 10 remaining subjects. Since sympathetic burst frequency remained constant over time in these 9 subjects, the mean value for MSNA in the 10th subject at 90 minutes was estimated from the average of his preceding burst frequency and burst incidence. Overall, the saline infusion had no significant effect on sympathetic nerve activity (Fig 2⇓) and plasma norepinephrine, epinephrine, plasma ET-1 (P=.052), plasma ANF, or cGMP concentrations (Table 2⇓), and HRV was stable over time (Table 3⇓).
Infusion of ANF increased its plasma concentration from 10±5 to 40±20 pmol/L (P<.001) and that of cGMP from 3.3±1.1 to 9.5±2.9 nmol/L (P<.0001) (Table 2⇑). These increases were evident within 30 minutes of infusion and coincided with the nadir of DBP. Subsequently, CVP was significantly reduced (Fig 2⇑). Systemic hemodynamics were otherwise unchanged (Fig 1⇑, Table 1⇑). Both FBF and FVC fell during ANF infusion (Table 1⇑). Reductions in CVP and DBP were accomplished by a significant increase in plasma epinephrine, but HR, MSNA, and plasma norepinephrine did not change (Table 2⇑, Fig 2⇑), and HRV was stable over time (Table 3⇑). Plasma ET-1 concentrations were identical at baseline and after 90 minutes of ANF infusion (Fig 1⇑).
Compared with saline ANF caused significantly greater reductions in CVP at 90 minutes (P<.04), in HR at 60 (P<.04) and 90 (P<.02) minutes, in cardiac output at 60 (P<.03) and 90 (P<.03) minutes, and in FBF at all three time periods (P<.05). FVR increased, with its peak at 60 minutes (P<.05).
Plasma ANF concentrations and cGMP levels tended to rise after ingestion of candoxatril, but at no time were these increases significant (Table 2⇑). Candoxatril caused a modest but significant rise in SBP at 90 minutes, along with a significant increase in plasma ET-1 concentrations (from 4.6±1.1 to 6.8±3.2 pmol/L; P<.02) (Fig 1⇑) and a significant reduction in CVP, first evident after 60 minutes; but it had no other effects on systemic or forearm hemodynamics (Table 1⇑). As with ANF, candoxatril caused no change in HR, MSNA, or plasma norepinephrine, but plasma epinephrine increased significantly (Table 2⇑). HRV was stable over time (Table 3⇑). The principal difference between these variables on the candoxatril and saline days was the absence of a rise in HR after the inhibitor (P<.03 at 30 and 60 minutes). ANF caused significantly greater reductions in CVP (P<.04 at 90 minutes), mean arterial pressure (P<.03 at 90 minutes), and FBF (P<.04 at 60 minutes) than candoxatril, but unlike candoxatril ANF did not affect plasma ET-1 concentrations.
The aim of this experiment was to determine the effects of increasing plasma ANF concentrations in healthy subjects, whether by its infusion or by reducing its metabolism by an orally active NEP inhibitor, on cardiac and systemic hemodynamics and on direct and indirect indices of sympathetic nervous system activity. Both candoxatril and low-dose ANF infusion had modest and relatively parallel short-term effects on hemodynamics. As anticipated, both interventions lowered CVP (Fig 2⇑). This reduction in preload occurred in the absence of any detectable increases in FVC and thus could not be attributed to venodilation. This decrease did not elicit any reflex increase in MSNA (Fig 2⇑) or plasma norepinephrine concentration or in the spectral representation of cardiac noradrenergic drive, but both ANF and candoxatril increased plasma epinephrine concentration. In contrast to the saline infusion, HR and stroke volume were not increased by either ANF or candoxatril. These observations therefore suggest that both interventions may specifically inhibit sympathetic nerve traffic to the heart and skeletal muscle. Whereas ANF caused a transient reduction in DBP, SBP increased gradually after candoxatril (+4 mm Hg), reaching significance 90 minutes after candoxatril administration. This increase cannot be attributed to a rise in cardiac output or activation of sympathetic outflow to resistance vessels in the calf. On the other hand, a significant increase in plasma ET-1 concentration was also observed at this time (Fig 1⇑). Effects on forearm hemodynamics were also dissimilar. ANF lowered FBF and FVC, whereas candoxatril had no significant effect on either of these variables.
Our discussion will focus on (1) relating these effects of a low-dose ANF infusion to our previous experimental observations, (2) similarities and differences in responses to ANF and candoxatril, (3) potential methodological limitations, and (4) evidence for a sympathetic modulatory effect of ANF at physiological and supraphysiological plasma concentrations.
Responses to ANF Infusion
Although ANF has natriuretic and diuretic effects in humans at physiological plasma concentrations,17 in most studies documenting an interaction with the autonomic nervous system plasma ANF concentrations were increased to supraphysiological levels. In our previous experiments consistent reductions in DBP (5 to 7 mm Hg), CVP (2 to 3 mm Hg), and systemic vascular resistance (approximately 25%) as well as consistent increases in HR, stroke volume, and cardiac output were observed when ANF was infused at much higher doses (50 μg followed by 50 ng/kg per minute, or 16 nmol followed by 16 pmol/kg per minute) over a shorter duration to achieve plasma concentrations in the range of 150 pmol/L.1 In those experiments we also demonstrated a relative inhibitory effect of ANF on sympathetic outflow, in that reductions in DBP and CVP were not accompanied by the reflex increases in MSNA and plasma norepinephrine concentrations elicited by similar hemodynamic changes during nitroprusside infusion, and sympathoneural responses to both the application and cessation of nonhypotensive lower-body negative pressure were significantly attenuated when compared with responses observed during nitroglycerin, infused as a hemodynamic control for ANF.1 Subsequent experiments have demonstrated an inhibitory effect of ANF on sympathetic neurotransmission at the ganglionic level.2 Frequency domain analysis of HRV has documented reductions in total spectral power and that component of harmonic spectral power commonly attributed to cardiac noradrenergic drive (ratio of low- to high-frequency power), without significantly affecting the rates of high-frequency to total power (parasympathetic indicator).15 These observations in the frequency domain provide further support for the concept that ANF acts on the autonomic nervous system to decrease sympathetic outflow.1 2 15 18 19 20
In the context of the present experiments, several of these previous observations—in particular, its effects on HR, HRV, stroke volume, and systemic vascular resistance—can now be considered a consequence of the supraphysiological ANF plasma concentrations achieved. However, in the present series there was a transient drop in DBP and a significant and sustained reduction in CVP, not accompanied by reflex sympathetic neural activation, indicating that those particular actions of ANF are manifest at physiological plasma concentrations.17
The forearm hemodynamic data differ from previous experiments by one of us (S.A.) and suggest that there may be qualitatively different responses to low- and high-dose systemic infusion of ANF in this vascular bed. We propose two possible explanations. When infused directly into the brachial artery of healthy young men in a high dose calculated to double baseline FBF (mean, 240±34 pmol/min), ANF caused significant increases in forearm venous distensibility and capillary filtration.12 A potential limitation of the plethysmographic method of assessing venous tone is that it infers changes in intravascular volume from changes in forearm circumference and assumes that any changes in extravascular volume caused by greater capillary filtration are trivial when these modest occluding pressures are applied for brief periods.11 It is not known whether increases in capillary filtration that could influence forearm hemodynamics occur at the much lower local forearm ANF concentration achieved in the present experiments. If so, subsequent changes in extravasated volume might contribute to the reduction in FVC 90 minutes into the ANF infusion. In recent experiments by Jansen et al21 intra-arterial infusion of ANF (3 pmol/min per 100 mL forearm volume) potentiated the forearm vasoconstrictor response to neurally released norepinephrine without augmenting forearm norepinephrine spillover into plasma or postjunctional responsiveness to coinfusion of norepinephrine with ANF. These authors therefore postulated the existence of a local mechanism by which ANF augments vasoconstrictor responses to α-adrenergic sympathetic stimulation. Such a mechanism could contribute to the reduction in FBF and FVC observed during ANF infusion, and the lower ANF concentrations observed after candoxatril may explain why forearm hemodynamics did not change when the atriopeptidase inhibitor was given.
Responses to Oral Candoxatril
ANF is rapidly inactivated in vivo by the NEP EC 18.104.22.168, located principally in the brush border membrane of the renal proximal tubule.8 22 Candoxatril is one of several atriopeptidase inhibitors developed for clinical use in hypertension.3 4 5 6 7 8 9 23 These compounds may also lower blood pressure by raising plasma concentrations of bradykinin and vasoactive peptides that are also metabolized by NEP8 22 (Fig 3⇓). Although the natriuretic action of candoxatril is mediated through the ANP receptor,24 a significant component of the fall in blood pressure after long-term NEP inhibition occurs independently of changes in plasma ANF concentrations.7 25
The principal limitation of the present study is that we did not achieve our goal of similar plasma levels of ANF on the two active intervention days. The dose and timing of candoxatril were selected on the basis of published data indicating significant increases in plasma ANF within 60 to 90 minutes of its oral ingestion.3 5 6 26 We observed slight increases in ANF and its second messenger cGMP, but they were not significant. This might explain why candoxatril did not lower DBP but cannot account for its effect on SBP. Moreover, candoxatril was active in these experiments, causing for example a significant reduction in CVP within 60 minutes; because many of its effects were similar to those observed during ANF, such activity was probably effected via ANF-related pathways. As with ANF, this decrease in CVP was not accompanied by reflex increases in MSNA or plasma norepinephrine, as might be expected if CVP fell by nonatriopeptidergic mechanisms. It is important to note that the natriuretic response to a single oral dose of candoxatril can occur independently of any changes in blood pressure,5 6 26 and in one study of subjects with essential hypertension, long-term (2 weeks) treatment with candoxatril lowered SBP and increased urinary but not plasma ANF concentrations.7
In a recent study the active metabolite candoxatrilat also caused a significant increase in SBP in healthy subjects.9 Because this rise was prevented by enalapril, it was suggested that it was caused by potentiation of angiotensin II, another substrate for NEP (Fig 3⇑). However, NEP in the renal brush border also appears to play a role in the metabolism of endothelin.8 27 28 29 When infused into the brachial artery, both endothelin and the NEP inhibitor thiorphan cause an increase in FVR.30 The magnitude of the increase in SBP in the present study is within the range anticipated from experiments in which endothelin was infused to achieve plasma levels similar to those in our subjects.31 The parallel increase in plasma endothelin in our subjects therefore suggests that the rise in SBP after candoxatril but not after exogenous ANF may be related to altered endothelin metabolism by the endopeptidase inhibitor. Whether such nonatriopeptidergic, nonadrenergic effects of NEP inhibition are functionally important in the treatment of conditions such as congestive heart failure or hypertension merits assessment.
The increases in HR and stroke volume 60 minutes into the saline infusion were not expected. The amount of saline infused (<70 mL) was too trivial to account for these changes and was infused on all 3 days. One possible explanation is that these changes were due to cutaneous vasodilation over the course of these experiments. By contrast, these variables remained constant over time on the ANF and candoxatril days, and when compared with saline by two-way ANOVA, there were significant inhibitory effects of ANF and candoxatril on these variables.
Sympathoinhibitory Effects of ANF
The concept that ANF interacts with the autonomic nervous system to alter the reflex control of the circulation has been the subject of recent reviews.19 20 For the most part, the ANF dose administered in experiments that have demonstrated such interactions has achieved plasma concentrations well above the normal physiological range. In the present series CVP fell after candoxatril administration and during low-dose ANF infusion. When elicited by parenteral vasodilators or by lower-body negative pressure, similar reductions in CVP cause significant reflex increases in MSNA.1 The lack of increase in MSNA, plasma norepinephrine, or the “cardiac sympathetic” component of HRV in the present series provides further support for the concept that this peptide is an important modulator of circulatory homeostasis via the autonomic nervous system at plasma concentrations within the physiological range. Similar modulation of renal sympathetic nerve activity may facilitate natriuresis. Because plasma epinephrine (unlike norepinephrine) rose significantly in response to the two interventions, both ANF and candoxatril may selectively inhibit sympathetic nerve traffic to muscle and the heart. Experiments that evaluate the effects of these two interventions on sympathoneural responses to arterial or cardiopulmonary baroreceptor reflex unloading or activation would provide a more direct test of this hypothesis.
Selected Abbreviations and Acronyms
|ANF||=||atrial natriuretic factor|
|CVP||=||central venous pressure|
|DBP||=||diastolic blood pressure|
|FBF||=||forearm blood flow|
|FVC||=||forearm venous compliance|
|FVR||=||forearm vascular resistance|
|HRV||=||heart rate variability|
|MSNA||=||muscle sympathetic nerve activity|
|SBP||=||systolic blood pressure|
This work was supported by Medical Research Council of Canada Operating Grant MT 9721 and by the 1994 International Society of Hypertension Pfizer Award. Dr Ando was supported by a Postgraduate Award from the Department of Medicine, University of Toronto, and by a stipend from this Medical Research Council Operating Grant. Drs M.A. Rahman and G.C. Butler held Fellowships from the Medical Research Council of Canada. The authors thank Pfizer Pharmaceuticals for their kind gift of candoxatril. Dr J.S. Floras is currently a Career Investigator of the Heart and Stroke Foundation of Ontario.
Reprint requests to John S. Floras, Division of Cardiology, Mount Sinai Hospital, Suite 1615, 600 University Ave, Toronto, Ontario, M5G 1X5, Canada.
- Received June 17, 1995.
- Revision received July 28, 1995.
- Accepted August 18, 1995.
Floras JS. Sympathoinhibitory effects of atrial natriuretic factor in normal humans. Circulation. 1990;81:1860-1873.
Floras JS. Inhibitory effects of atrial natriuretic factor on sympathetic ganglionic transmission in humans. Am J Physiol. 1995;269:R406-R412.
Singer DR, Markandu ND, Buckley MG, Miller MA, Sagnella GA, MacGregor GA. Dietary sodium and inhibition of neutral endopeptidase 24.11 in essential hypertension. Hypertension. 1991;18:798-804.
Tunny TJ, Ziesak MD, Armstrong R, Klemm SA, Stowasser M, Finn WL, Gordon RD. Inhibition of endopeptidase EC 22.214.171.124 by candoxatril lowered blood pressure and increased urinary but not plasma atrial natriuretic peptide in essential hypertension. J Hypertens. 1993;11(suppl):S222-S223.
Motwani JG, Lang CC, Cramb G, Struthers AD. Natriuretic response to neutral endopeptidase inhibition is blunted by enalapril in healthy men. Hypertension. 1995;25:637-642.
Naughton M, Rahman MA, Hara K, Floras JS, Bradley TD. Effect of continuous positive airway pressure on intrathoracic and left ventricular transmural pressure in patients with congestive heart failure. Circulation. 1995;91:1725-1731.
Manyari DE, Malkinson TJ, Robinson V, Smith ER, Cooper KE. Acute changes in forearm venous volume and tone using radionuclide plethysmography. Am J Physiol. 1988;255:H947-H952.
Yamamoto Y, Hughson RL. Coarse-graining spectral analysis: new method for studying heart rate variability. J Appl Physiol. 1991;71:1143-1150.
Butler GC, Yamamoto Y, Xing HC, Northey DR, Hughson RL. Heart rate variability and fractal dimension during orthostatic challenges. J Appl Physiol. 1993;75:2602-2612.
Butler GC, Senn BL, Floras JS. Influence of atrial natriuretic factor on heart rate variability in normal man. Am J Physiol. 1994;267:H500-H505.
Campbell PJ, Skorecki KL, Logan AG, Wong PY, Leung WM, Greig P, Blendis LM. Acute effects of peritoneovenous shunting on plasma atrial natriuretic peptide in cirrhotic patients with massive refractory ascites. Am J Med. 1988;84:112-119.
Volpe M. Atrial natriuretic peptide and the baroreflex control of circulation. Am J Hypertens. 1992;5:488-493.
Erdos EG, Skidgel RA. Neutral endopeptidase 24.11 (enkephalinase) and related regulators of peptide hormones. FASEB J. 1989;3:145-151.
Ogihara T, Rakugi H, Matsuo K, Yu H, Nagano M, Mikami H. Antihypertensive effects of the neutral endopeptidase inhibitor SCH 42495 in essential hypertension. Am J Hypertens. 1994;7:943-947.
Richards M, Espiner E, Frampton C, Ikram H, Yandel T, Sopwith M, Cussans N. Inhibition of endopeptidase EC 24.11 in humans: renal and endocrine effects. Hypertension. 1990;16:269-276.
Abassi ZA, Tate JE, Golomb E, Keiser HR. Role of neutral endopeptidase in the metabolism of endothelin. Hypertension. 1992;20:89-95.
Fujita K, Matsumura Y, Kita S, Hisaki K, Takaoka M, Morimoto S. Phosphoramidon-sensitive conversion of big endothelin-1 and degradation of endothelin-1 in rat kidney. Hypertension. 1994;24:227-233.
Kaasjager KAH, Koomans HA, Rabelink TJ. Effectiveness of enalapril versus nifedipine to antagonize blood pressure and the renal response to endothelin in humans. Hypertension. 1995;25:620-625.