Author + information
- Received July 8, 2005
- Revision received February 23, 2006
- Accepted April 4, 2006
- Published online September 19, 2006.
- Rebecca H. Ritchie, BSc(Hons), PhD⁎,†,1,
- Christopher J. Zeitz, MBBS, PhD, FRACP⁎,2,
- Ronald D. Wuttke, BSc⁎,
- John T.Y. Hii, BMBS, FRACP⁎ and
- John D. Horowitz, MBBS, PhD, FRACP⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. John D. Horowitz, Cardiology Unit, The Queen Elizabeth Hospital, 28 Woodville Road, Woodville, SA 5011, Australia.
Objectives This study sought to compare the influence of changes in systolic interval on the negative inotropic effects of metoprolol, sotalol, and verapamil in patients with ischemic heart disease.
Background The long-term symptomatic and prognostic effects of antiarrhythmic agents are not easily predicted on the basis of acute hemodynamic actions at rest, but may be unmasked during tachycardia. However, this has not been studied extensively in vivo.
Methods The force-interval relationship of the intact human left ventricle was examined before and 10 min after intravenous bolus administration of the negatively inotropic agents metoprolol, sotalol, or verapamil in patients undergoing diagnostic cardiac catheterization.
Results All three drugs similarly reduced maximal rate of increase of left ventricular pressures (LV+dP/dtmax) by approximately 10%, but diversely modified the force-interval relationship. The parameter c (the reduction in LV+dP/dtmaxof a fixed premature stimulus on mechanical restitution) was significantly reduced by metoprolol (by 9 ± 4%, p < 0.05), was increased by verapamil (by 6 ± 2%, p < 0.05), and was not significantly changed by sotalol. Similarly, metoprolol had a minimal effect on the extent of frequency potentiation, whereas sotalol and verapamil attenuated frequency potentiation (the relative response to 10 s of rapid pacing was 1.19 ± 0.58-fold, 0.07 ± 0.35-fold, and 0.03 ± 0.17-fold of the baseline response after 10 min of metoprolol, sotalol, or verapamil, respectively).
Conclusions These results show that the negative inotropic effects of metoprolol are attenuated and those of verapamil are accentuated at short cycle lengths; sotalol is intermediate between the two. These properties may contribute to the relative safety of these agents in patients prone to hemodynamic deterioration during sustained tachycardia.
Onset of tachycardia in patients administered class I (sodium-channel blocker) or class IV (calcium-channel antagonist) antiarrhythmic drugs is associated with increased risk of acute hemodynamic collapse, despite these agents being well tolerated in sinus rhythm (1,2). Changes in heart rate may thus unmask negative inotropic effects of cardioactive drugs not apparent at rest, which may contribute to increased mortality during treatment with such agents in patients with impaired left ventricular systolic function (3,4). Conversely, class II antiarrhythmic agents, beta-adrenoceptor antagonists, are not poorly tolerated at faster heart rates (5), suggesting differential modulation of the force-frequency relationship by antiarrhythmic agents.
The relationship between alterations in stimulation rate and myocardial contractile performance can be studied using either frequency potentiation and/or mechanical restitution. Frequency potentiation, also known as the staircase or Treppe phenomenon, is commonly used in vitro (6,7), and is illustrated by the incremental reductions in calcium-channel current and contractile force induced by verapamil in vitro with progressive increases in stimulation frequency (8,9). However, the sustained tachycardia of frequency potentiation in vivo may result in neurohumoral activation, changed loading conditions, ischemia, or even hemodynamic deterioration (10–13). Such effects could distort the force-interval relationship, and thus limit utility of frequency potentiation.
The force-interval relationship can also be examined (both in vitro and in vivo) using mechanical restitution curves (MRC). Mechanical restitution is the recovery of myocardial contractility after a non–steady-state beat (7). We have recently described a mathematical model of the MRC that has proven highly reproducible in patients under investigation for ischemic heart disease (14,15). In the current investigation, we studied the acute effects of 3 pharmacologically different antiarrhythmic agents on the force-interval relationship, metoprolol (a selective β1-adrenoceptor antagonist), d,l-sotalol (a nonselective β-adrenoceptor antagonist with additional outward delayed rectifier potassium current blocking properties) and verapamil (an L-type calcium-channel antagonist). All three drugs elicit negative inotropic effects in sinus rhythm. Using both MRC construction and frequency potentiation analysis, we tested the hypothesis that the negative inotropic effects of metoprolol, but not those of verapamil, were independent of changes in R-R interval in patients with ischemic heart disease. Although previous data have suggested that the negative inotropic effects of verapamil might be accentuated at higher stimulation frequencies (9,10), clinically based hemodynamic comparisons of these agents have been lacking.
Patients with stable symptoms were selected from those undergoing routine diagnostic cardiac catheterization and coronary arteriography for the investigation of chest pain. Exclusion criteria included unstable angina pectoris, significant left main coronary artery stenosis, electrocardiographic (ECG) evidence of abnormal conduction intervals, clinically significant valvular disease, recent myocardial infarction (in the last 3 months), and severe impairment of left ventricular systolic function (ejection fraction <30%), in addition to clinically significant renal or hepatic disease. The protocol was approved by the Ethics of Human Research Committee of The Queen Elizabeth Hospital, and prior informed consent was obtained.
Administration of all beta-adrenoceptor and calcium-channel antagonists was ceased at least 5 half-lives before the study. Oral diazepam and diphenhydramine were administered approximately 30 min before cardiac catheterization as premedication. Right and left cardiac catheterization and coronary arteriography were performed under local anesthesia (1% lidocaine) using the Judkins approach via femoral arterial (16) and venous sheaths. The research procedure commenced at the end of the routine catheterization. A bipolar pacing lead was positioned in the right atrium, and a 4-F micromanometer-tipped catheter (Millar Instruments, Houston, Texas) was inserted via the femoral artery sheath into the left ventricle for measurement of left ventricular pressure and maximal rate of increase of left ventricular pressures (LV+dP/dtmax). A 7-F Swan-Ganz catheter was positioned in the pulmonary artery for determination of cardiac output via thermodilution. Incremental radiation exposure associated with the research procedure was minimal, limited to screening to check Millar catheter position.
Patients underwent continuous baseline atrial pacing at a rate 7 ± 1% above spontaneous heart rate to maintain a constant heart rate throughout the procedure. Cardiac output (the average of at least three readings at each time point), mean arterial pressure (MAP), ECG parameters, and left ventricular pressure and its first derivative LV+dP/dtmaxwere recorded continuously. After acquisition of baseline data, including MRC and frequency potentiation determination, patients were sequentially allocated to receive metoprolol (4 mg; n = 15), sotalol (20 mg; n = 15), or verapamil (4 mg; n = 17) as a rapid intravenous bolus. Sodium nitroprusside was also investigated in an additional subgroup of patients (10 to 20 μg/min intravenous infusion until approximately a 10% reduction in MAP was observed, n = 5) to determine the influence, in isolation, of altered loading conditions on the MRC. The dose of each agent was chosen from the lower end of the dosage range used clinically. Hemodynamic measurements and serial MRCs were obtained at frequent intervals up to 10 min after administration. Frequency potentiation was re-examined at 10 min.
The MRCs were constructed as previously described (14,15). Briefly, premature stimuli of twice the threshold strength were inserted every 8 beats during baseline pacing, at progressively shorter R-R intervals, until refractoriness prevented impulse conduction. The contractile strength of each beat, LV+dP/dtmax, was plotted as a function of the R-R interval. The LV+dP/dtmaxwas expressed as a percent of that observed in the drug-free state during baseline pacing, and R-R interval as a percent of the cycle length of the baseline-pacing rate. This was then fitted to one-half of a rectangular hyperbolic function, described by:where x is the R-R interval of the premature stimulus and y is LV+dP/dtmax. The horizontal and vertical asymptotes are represented by a and d, respectively. The parameter c, the difference between the calculated values of LV+dP/dtmaxfrom the fitted curve when the R-R interval is 100%, and 60% (illustrated in Fig. 1),describes contractile sensitivity to reductions in R-R interval for each patient (14). We then calculated the change in the parameter c 10 min after treatment by subtracting it from the pretreatment value, to obtain Δ-parameter c. Thus, an increase in c after drug administration (i.e., positive Δ-parameter c) would indicate directly rate-dependent negative inotropic effects.
The force-interval relationship was also assessed using frequency potentiation by determining the influence of 1 min of rapid pacing (34 ± 2% above the rate of baseline pacing) on LV+dP/dtmaxin 10-s intervals. This protocol was not performed in patients with severe angina pectoris, and was precluded by the development of atrioventricular block in some patients. We then further analyzed the frequency potentiation response by expressing the pacing-induced increase in LV+dP/dtmaxafter drug administration in each patient as a ratio of the response observed before treatment for each of metoprolol, sotalol, and verapamil to calculate the relative frequency potentiation response.
Results were expressed as mean ± SE. The goodness-of-fit of the MRC model for each individual patient was determined using residual standard deviations. Confidence intervals (Cis) at 95% were derived for these standard deviations. One-factor analysis of variance was used to compare patient characteristics at baseline in the four groups, and with repeated measures analysis (Dunnett), to examine the time course of LV+dP/dtmaxand c as appropriate. Paired ttests were used to compare the hemodynamic and ECG effects of the drugs studied, before and 10 min after injection. The differential effects of the three agents, metoprolol, verapamil, and sotalol, on MRC analysis were examined via one-way analysis of variance of Δ-parameter c across all three patient groups. To determine whether the 3 agents attenuated the frequency potentiation response, we compared the relative frequency potentiation response across all three patient groups using a 1-way analysis of variance. Statistical significance was accepted at the p < 0.05 level.
The clinical characteristics before drug injection of all 52 patients studied are summarized in Table 1.The groups were generally well matched, with predominantly normal left ventricular systolic function, and no significant differences in characteristics between the groups. Clinically significant coronary artery disease (stenosis ≥50% in at least 1 major branch of a coronary artery) was present in 36 patients. The procedure was well tolerated in all patients.
Hemodynamic effects during baseline pacing
The doses for each of metoprolol, sotalol, and verapamil used in the present study were chosen with the intention of achieving similar negative inotropic effects at baseline pacing: the three negatively inotropic drugs induced approximately a 10% reduction in LV+dP/dtmax(Fig. 2).Table 2summarizes the hemodynamic and ECG effects of all drugs studied during baseline pacing, at the time of peak effect (10 min after drug injection for metoprolol, sotalol, and verapamil, and at the time of maximal hypotensive effect for sodium nitroprusside). The MAP was significantly reduced by both verapamil and sodium nitroprusside (as a requirement of the length of the infusion period), but not by either metoprolol or sotalol. Sodium nitroprusside had no additional hemodynamic effects, although pulmonary capillary wedge pressure also tended to decrease (from 13 ± 2 mm Hg to 10 ± 1 mm Hg, p = NS).
The influence of metoprolol, sotalol, and verapamil on the MRC is shown in Figure 3before and 10 min after injection (the time of peak effect). The residual SDs of the model (the measure of goodness-of-fit) for these drugs were 4.9 (95% CI 3.3 to 6.4), 5.1 (95% CI 3.5 to 6.7), and 8.3 (95% CI 5.9 to 10.6; p < 0.05 vs. metoprolol and verapamil) for metoprolol, verapamil, and sotalol, respectively. Although all three drugs exerted a similar negative inotropic effect during baseline pacing (Fig. 2), these effects differed on MRC analysis. Metoprolol decreased LV+dP/dtmaxby 12 ± 3% during baseline pacing. However, this negative inotropic effect became less marked at shorter R-R intervals, and was completely abolished when the R-R interval was reduced to 60%. On the fitted MRC, LV+dP/dtmaxwas 72 ± 4% and 70 ± 3% at an R-R interval of 60% before and after metoprolol, respectively (Fig. 3A). This was reflected by a significant reduction in c from 28 ± 4% to 21 ± 3% (Fig. 4A;p < 0.05). In the group of patients allocated to receive sotalol, LV+dP/dtmaxwas reduced by 11 ± 3% during baseline pacing. On MRC analysis, the negative inotropic effect of sotalol was virtually unchanged as the R-R interval decreased. The LV+dP/dtmaxon the fitted MRC was 72 ± 6% and 59 ± 7% at an R-R interval of 60% before and after sotalol, respectively (Fig. 3B), and c did not fluctuate significantly (from 29 ± 6% to 31 ± 7%, p = NS) (Fig. 4B). In contrast with metoprolol and sotalol, verapamil tended to induce a divergence of the MRC, with the negative inotropic effect (8 ± 2% reduction in LV+dP/dtmaxat baseline) progressively accentuated as the R-R interval decreased. On the fitted MRC, LV+dP/dtmaxwas 81 ± 4% and 67 ± 5% at an R-R interval of 60% before and after verapamil, respectively (Fig. 3C). This corresponded to an increase in c from 18 ± 4% at baseline to 24 ± 4%, indicative of a rate-dependent effect (p < 0.05) (Fig. 4C). Moreover, when Δ-parameter c after 10 min was compared among metoprolol, sotalol, and verapamil, the MRC response to metoprolol was significantly different than that obtained with verapamil (p < 0.05) (Fig. 4B). The significant decrease in MAP induced by sodium nitroprusside, however, was not accompanied by significant alterations in LV+dP/dtmaxeither during baseline pacing (−5 ± 4% vs. baseline, p = NS) or at shorter R-R intervals. No significant fluctuations in the parameter c were observed (from 43 ± 6% to 43 ± 7%, p = NS, results not shown).
The force-interval relationship was also investigated using frequency potentiation analysis before and 10 min after administration of each of the three negatively inotropic drugs. Figure 5shows the time course of LV+dP/dtmaxduring rapid pacing. Before drug administration, the effects of rapid pacing were comparable in the 3 treatment groups, showing an initial increase of approximately 20% in LV+dP/dtmax, which was preserved for the full minute. As with MRC analysis, all three agents examined showed different effects on the extent of frequency potentiation (Fig. 5), especially evident after 10 s of rapid pacing. Metoprolol did not affect the frequency potentiation response: the pacing-induced increases in LV+dP/dtmaxbefore and after the drug were maintained (Fig. 5A). This was not the case for verapamil: the pacing-induced increase in LV+dP/dtmaxseemed to be markedly suppressed during the 1 min of rapid pacing (Fig. 5C). The results for sotalol were intermediate between those of metoprolol and verapamil (Fig. 5C). As shown in Figure 5D, the relative frequency potentiation response after metoprolol was unchanged after 10 s of rapid pacing; conversely, both sotalol and verapamil significantly attenuated the relative frequency potentiation response (p < 0.005). After 60 s of rapid pacing (at which time ischemia and/or neurohumoral activation may be evident [10,12]), sotalol, however, no longer the attenuated the frequency potentiation response (results not shown).
The results of the current study show that the negative inotropic effects during tachycardia of a pharmacologically heterogeneous group of antiarrhythmic drugs cannot be predicted on the basis of their observed effects in the resting state. Metoprolol, sotalol and verapamil were administered acutely at doses that elicited comparable negative inotropic effects at rest (Fig. 2) and minimal changes in pulmonary capillary wedge pressure. However, these three agents exerted markedly different negative inotropic effects at shorter R-R intervals, as determined from both MRC (Figs. 3 and 4) and frequency potentiation analysis (Fig. 5). The negative inotropic effects of metoprolol were markedly attenuated at short cycle lengths on MRC analysis, and metoprolol did not attenuate the inotropic response associated with frequency potentiation. Conversely, verapamil exerted directly rate-dependent negative inotropic effects, virtually ablating the frequency potentiation response; accentuation of verapamil’s negative inotropy was also evident at short R-R intervals on MRC analysis. The effects of sotalol were intermediate between those of metoprolol and verapamil, relatively independent of R-R interval on MRC analysis and some attenuation of the frequency potentiation response.
Construction of MRCs was the primary methodology used in the current study for the quantitative examination of drug effects on the force-interval relationship. We have previously shown that this is highly reproducible, with significant determinants of the parameter c including R-R interval (held constant in the present investigation) and left ventricular ejection fraction (14,15). Sodium nitroprusside, used in part because of the known insensitivity of the force-frequency relationship to nitric oxide (17), exerted no significant effects on mechanical restitution, suggesting that the parameter c is relatively independent of small changes in preload and afterload. Nevertheless, it remains possible that differential effects on preload might have influenced some of the observed differences between verapamil and metoprolol/sotalol. The use of additional measures of the inotropic state, and in particular the simultaneous measurement of left ventricular pressure and volume, would have provided more load-independent assessment of drug effect. The reductions in LV+dP/dtmaxinduced by metoprolol and sotalol in the present study are consistent with previous investigations in humans (18–21). However, we now show that the negative inotropic effect of metoprolol is reverse rate-dependent on MRC analysis. These data are consistent with recent suggestions that metoprolol normalizes the ventricular force-frequency relationship in patients with heart failure (5). Conversely, sotalol did not significantly influence MRC. Also on MRC analysis, we quantitatively showed that the negative inotropic effect of verapamil was accentuated at shorter R-R intervals in humans in vivo. The rate-dependent nature of this negative inotropic effect has been clearly defined in animal models in vitro (8,9). However, no evidence of this type for verapamil or any other calcium antagonist in humans was previously available. Clinical experience with the acute administration of verapamil has suggested the potential for acute hypotensive and negative inotropic sequelae (2,4), a complication that is rarely observed in the absence of tachyarrhythmias.
The most appropriate examination of the effects of all three agents on frequency potentiation is after 10 s of rapid pacing (Fig. 5D) because more prolonged tachycardia may induce ischemia (12) and/or neurohumoral activation (10) in some patients. Nevertheless, the three agents examined disparate effects on frequency potentiation irrespective of the period of rapid pacing considered. Metoprolol had no detectable effect, whereas verapamil significantly (and sotalol to a lesser extent) impaired the frequency potentiation response. These disparate effects of the agents examined may reflect the underlying physiological changes involved in mechanical restitution and frequency potentiation. Frequency potentiation in an intact circulation represents a complex model comprising both negative (mechanical restitution-like) and positive inotropic (Treppe-like) components. The cellular correlates of this model cannot be ascertained in the current setting. Nevertheless, the rank order of rate-related negative inotropic effects (verapamil > sotalol > metoprolol) was identical to that seen with MRC analysis. The incomplete MRC, the principal measure of the force-interval relationship used in the present study, reflects availability of calcium for contractile performance. Conventionally it has been regarded that a major determinant of mechanical restitution at low stimulation frequencies is calcium release from the sarcoplasmic reticulum, whereas trans-sarcolemmal calcium flux becomes increasingly important at high stimulation frequencies (8,9,22,23). The diverse effects of metoprolol and verapamil at shorter R-R intervals observed in the current study may simply reflect different sites of action on frequency-dependent cellular physiology. The interaction between verapamil and the L-type calcium channel per se is frequency-dependent: the drug binds preferentially to its receptor when the channel is in the open state (8). No analogous cellular phenomenon has been observed for beta-adrenoceptor ligands. The major finding of attenuation of the negative inotropic effect of metoprolol at short cycle lengths is of great interest. Superficially, this is a somewhat paradoxical observation because previous studies in a range of myocardial preparations (24) have suggested that the positive inotropic response to beta-adrenoceptor agonists increases with stimulation frequency. Available data for beta-adrenoceptor antagonists, however, are consistent with our observations: both chronic therapy of heart failure patients with metoprolol and acute propranolol treatment in intact canine ventricles are associated with attenuated negative inotropic effects at increased heart rates (5,25).
The effects of sotalol on MRC and frequency potentiation clearly differed from those of metoprolol in the present study. The only previously reported investigation of sotalol effects in human ventricular myocardium (26) failed to show any effect of either d,l-sotalol or of d-sotalol (which lacks effects at the beta-adrenoceptor while retaining potassium-channel blockade), but the potential for rate-related inotropic interactions was not examined. The results of the current study imply that the potassium-channel blocking actions of sotalol modulate its inotropic effects, resulting in incremental negative inotropy during tachycardia.
This was essentially a study of patients with intact left ventricular systolic function (by nature of the selection criteria). It is possible that studying patients with impaired systolic function would produce different results with respect to the relationship between MRC and frequency potentiation data: such patients might show an underlying defect of intracellular calcium availability (27,28), the consequences of which would increase progressively as heart rate increased (29). Furthermore, we did not investigate the mechanisms underlying the attenuation of frequency potentiation effects (attributed to progressive induction of ischemia and counter-regulatory neurohumoral activation). The use of a pacing regimen that was sufficiently brief as to be associated with no overt ischemia in any patient clearly does not preclude the possibility that some patients might have developed subclinical ischemia during the latter stages of the frequency potentiation protocol. On the other hand, it is unlikely that the abolition of bradycardia-related anti-ischemic effects (in the cases especially of metoprolol and sotalol) would have differentially influenced the current results. Lastly, the potential for a correlation between changes in the force-interval relationship and simultaneous evaluation of cellular electrophysiology, difficult to accurately obtain in vivo, was not sought. Nevertheless, such information might have provided additional insight, particularly in the case of sotalol.
In summary, we have shown that the negative inotropic effects of metoprolol are attenuated at short cycle lengths. This would in theory be a salutary effect as regards maintenance of hemodynamic status during tachycardia, especially in the presence of impaired ventricular function, but might limit the effectiveness of the drug as an antianginal agent. Conversely, verapamil shows the potential for hemodynamic deterioration during tachycardia in susceptible individuals. This may limit its utility in patients with angina plus left ventricular dysfunction (30). Heterogeneity between effects of metoprolol and sotalol may also be important, illustrated by the deleterious effects of d-sotalol (31). Furthermore, the methodology used in this study might well prove useful both during preclinical studies and in phase 1 human investigations of new cardioactive agents to identify potential for hemodynamic deterioration during tachycardia.
The authors acknowledge the assistance of Professor Richard Jarrett, Department of Statistics, The University of Adelaide. The authors thank the staff of the Cardiac Catheterization Laboratory, The Queen Elizabeth Hospital, for assistance with the experimental procedure, and Mr. J. Pearce and Mr. B. Braysher for assistance with the mathematical mechanical restitution curve model.
↵1 Dr. Ritchie was a University of Adelaide (Adelaide, SA, Australia) and Queen Elizabeth Hospital Research Foundation (Woodville, SA, Australia) Postgraduate Scholar.
↵2 Dr. Zeitz was a National Heart Foundation Postgraduate Scholar.
Supported by grants from the National Heart Foundation (Deakin, ACT, Australia) and the Merck Foundation (South Granville, NSW, Australia).
- Abbreviations and Acronyms
- confidence interval
- maximal rate of increase of left ventricular pressures
- mean arterial pressure
- mechanical restitution curve
- Received July 8, 2005.
- Revision received February 23, 2006.
- Accepted April 4, 2006.
- American College of Cardiology Foundation
- Andersson B.,
- Stromblad S.O.,
- Lomsky M.,
- Waagstein F.
- Koch-Weser J.,
- Blinks J.R.
- Seed W.A.,
- Walker J.M.
- Borzak S.,
- Murphy S.,
- Marsh J.D.
- Aroesty J.M.,
- McKay R.G.,
- Heller G.V.,
- et al.
- Linhart J.W.,
- Hildner F.J.,
- Barold S.S.,
- Lister J.W.,
- Samet P.
- Cotton J.M.,
- Kearney M.T.,
- MacCarthy P.A.,
- et al.
- Hutton I.,
- Lorimer A.R.,
- Hillis W.S.,
- et al.
- Cooper I.C.,
- Fry C.H.
- Holubarsch C.,
- Schneider R.,
- Pieske B.,
- et al.
- Hasenfuss G.
- Richard S.,
- Leclercq F.,
- Lemaire S.,
- Piot C.,
- Nargeot J.