Inhibition of the Pacemaker Current: A Bradycardic Therapy for Off-Pump Coronary Operations

Transcription

1 Inhibition of the Pacemaker Current: A Bradycardic Therapy for Off-Pump Coronary Operations Alain Bel, MD, Louis P. Perrault, MD, PhD, Bouchaib Faris, MD, Christian Mouas, Jean-Paul Vilaine, MD, and Philippe Menasché, MD, PhD Department of Cardiovascular Surgery and INSERM U-127, Hôpital Lariboisière, Paris, and the Servier Research Institute, Suresnes, France Background. The accurate performance of coronary anastomoses on the beating heart requires some form of myocardial immobilization that can be achieved pharmacologically. Different classes of drugs can be used to induce bradycardia, but the most effective in this setting of off-pump operation has not yet been determined. Methods. Fifty-six isolated buffer-perfused rabbit hearts were divided into seven equal groups. Control hearts were continuously perfused throughout the experimental time course. A second group of hearts underwent 60 minutes of potassium arrest (at 37 C) followed by 1 hour of reperfusion. The following pharmacologic approaches were tested in the remaining five groups: shortacting -blockade (esmolol, mol/l and mol/l), opening of adenosine triphosphate-dependent potassium channels (nicorandil, 10 3 mol/l and 10 5 mol/l), and inhibition of the pacemaker current, which largely accounts for the diastolic depolarization of sinoatrial node cells (S , mol/l). Each drug was infused at a constant rate for 60 minutes, after which hearts were perfused for 1 additional hour with drug-free buffer. Heart rate and isovolumic measurements of function and coronary flow were serially taken during and after drug infusion. Results. The worst recovery of systolic and, moreover, diastolic function was yielded by potassium arrest. Neither esmolol nor nicorandil was able to induce a significant bradycardia. However, nicorandil did not impair function which, conversely, was markedly depressed after esmolol therapy. Significant bradycardia (p < versus corresponding baseline values and versus all other groups) was only achieved with pacemaker current inhibition, which was otherwise associated with an excellent preservation of contractility, diastolic function, and coronary flow. Conclusions. Inhibition of the pacemaker current seems to be an effective approach for inducing intraoperative bradycardia without compromising left ventricular function or flow. (Ann Thorac Surg 1998;66:148 52) 1998 by The Society of Thoracic Surgeons Coronary artery bypass operations without cardiopulmonary bypass are currently receiving a renewed interest. The accurate performance of distal coronary anastomoses is clearly made more difficult by the beating state of the heart. This problem can be addressed by mechanical stabilization or bradycardic drugs. In this setting, the most commonly used approach is shortacting -blockade with esmolol [1 3]. However, there is little information on the chronotropic and inotropic effects of this drug when given in the context of off-pump operations. The present study was designed to address this issue by comparing short-acting -blockade with two newer approaches for slowing the heart rate: opening of adenosine triphosphate-dependent potassium channels and inhibition of the pacemaker current. Material and Methods A total of 56 rabbits were used in this study. All animals received humane care in compliance with the Guide for Accepted for publication March 5, Address reprint requests to Dr Menasché, Department of Cardiovascular Surgery, Hôpital Lariboisière, 2, rue Ambroise Paré, Paris, France. the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH publication 85-23, revised 1985). Experimental Preparation Hearts were obtained from New Zealand white rabbits weighing an average of 1 kg and anesthesized with ketamine and xylazine (both 5 mg/kg). After intravenous heparinization (100 IU/kg), the hearts were quickly excised and mounted on a nonrecirculating Langendorff column where they were perfused retrogradely at 100 cm H 2 O with filtered Krebs Henseleit solution (in millimoles per liter: NaCl, 118; KCl, 4.7; MgSO 4, 1.2; NaHCO 3, 25; KH 2 PO 4, 1.2; CaCl 2, 2.5; glucose, 11) bubbled with 95%O 2 /5%CO 2. Both the perfusion apparatus and the heart chamber were water jacketed so as to maintain myocardial temperature at 37 C throughout the experimental time course. A latex balloon filled with saline solution and connected to a pressure transducer (model P 23 ID; Gould, Cleveland, OH) was introduced into the left ventricle. The first derivative of left ventricular developed pressure (dp/dt) was obtained from a differentiator (model ; Gould). Heart rate and functional data were 1998 by The Society of Thoracic Surgeons /98/$19.00 Published by Elsevier Science Inc PII S (98)

2 Ann Thorac Surg BEL ET AL 1998;66: BRADYCARDIA DURING OFF-PUMP OPERATIONS 149 displayed on a Schlumberger model OM-4502 chart recorder (Enertec, St. Etienne, France). The coronary effluent was collected in a beaker, and the flow was determined volumetrically. Pacing was not used at any time throughout the experimental protocol. Experimental Protocol and Groups Hearts were allowed to stabilize for 20 minutes during which baseline measurements of heart rate and function were taken. These measurements were made after inflation of the balloon at the volume that yielded a left ventricular end-diastolic pressure of 10 mm Hg. This volume was subsequently kept constant during the remainder of the experiment. Hearts were then randomly assigned to one of the following seven groups (eight hearts per group). One group of hearts were continuously perfused with drugfree Krebs buffer for 140 minutes and served as controls. A second group of hearts underwent 60 minutes of normothermic potassium arrest (20 mmol/l added to the Krebs buffer) followed by 60 minutes of reperfusion. Hearts of the remaining five groups received one of the drugs under study. Each drug, dissolved in Krebs buffer, was infused over 60 minutes at a constant rate (1.6 ml/min) using an electrically driven syringe pump (Vial, Grenoble, France). This syringe was connected to a catheter, the distal end of which was attached by a three-way stopcock to the retrograde aortic cannula immediately above the heart where drugs mixed with the bulk perfusate. At the completion of drug delivery, hearts were perfused for an additional 60-minute period with drug-free Krebs buffer. Measurements of heart rate and isovolumic function and flow were taken at 15-minute intervals throughout the periods of drug infusion and postdrug perfusion (except for the potassium-arrested group in which no recordings were obviously made during arrest). The drugs under investigation were: The short-acting -blocker esmolol at two concentrations (high, mol/l; low, mol/l), the adenosine triphosphatedependent potassium channel opener nicorandil at two concentrations (high, 10 3 mol/l; low, 10 5 mol/l), and the pacemaker current inhibitor S ( mol/l). These drugs were provided by the respective manufacturers (esmolol; Gensia, Bracknell, UK; nicorandil; Rhône-Poulenc Rorer, Collegeville, PA; and S ; Servier Research Institute, Suresnes, France). The concentrations of esmolol, nicorandil, and S were selected from literature data pertaining to similar isolated heart models [4 8]. Statistics Results are expressed as mean standard error of the mean. Differences were determined by two-factor repeated-measures analysis of variance followed by Student s post hoc t tests. A p value less than 0.05 was considered statistically significant. Fig 1. Comparative effects of the drugs on heart rate. (A) Values at the end of the 60-minute period of drug infusion (*p 0.02 versus control, p 0.03 versus E high, p versus E low, p versus N high and low). (B) Pooled values over the 60- minute period of postdrug perfusion (*p versus all groups). Values are given as mean standard error of the mean. Each group consisted of eight hearts. (C control (untreated) group; K potassium arrest; E high esmolol high dose ( mol/l); E low esmolol low dose ( mol/l); N high nicorandil high dose (10 3 mol/l); N low nicorandil low dose (10 5 mol/l); S S ( mol/l).) Results Baseline values for all parameters were not significantly different among the seven groups. Pooled group averages were as follows: heart rate, beats/min; left ventricular diastolic pressure, mm Hg; first derivative of left ventricular pressure (dp/dt), 2, mm Hg/s 1 ; and coronary flow, 44 1 ml/min. The most pronounced bradycardic effect was achieved with the drug that inhibits the pacemaker current (S ). In this group, heart rate was significantly lower than in all other groups, both at the end of drug infusion and over the ensuing 60-minute period of drug-free buffer perfusion (Fig 1). Furthermore, analysis of variance demonstrated a significant (p 0.009) time-related effect that was then specified by posthoc paired t tests showing that heart rate values measured in S treated hearts after drug infusion and at the end of the 60-minute postdrug perfusion were significantly lower than the corresponding baseline values ( 24% and 22%, respectively, p ). Thus, resumption of drug-free buffer perfusion did not relieve bradycardia, which persisted over the 1-hour postdrug observation period. At the time of discontinuation of drug infusion, the highest diastolic pressures were seen in the two esmololtreated groups (Fig 2). This effect was sustained over time only for the high-dose esmolol treatment that, during the subsequent drug-free perfusion period, shared with potassium arrest the worst recovery of diastolic function.

3 150 BEL ET AL Ann Thorac Surg BRADYCARDIA DURING OFF-PUMP OPERATIONS 1998;66: Fig 2. Comparative effects of the drugs on left ventricular diastolic pressure. (A) Values at the end of the 60-minute period of drug infusion (*p versus all groups; #p 0.04 versus N high, p 0.02 versus N low, p versus S). (B) Pooled values over the 60-minute period of postdrug perfusion (*p versus all groups; #p versus all groups). Values are given as mean standard error of the mean. Each group consisted of eight hearts. (C control (untreated) group; K potassium arrest; E high esmolol high dose ( mol/l); E low esmolol low dose ( mol/l); N high nicorandil high dose (10 3 mol/l); N low nicorandil low dose (10 5 mol/l); S S ( mol/ L).) Both during and after drug infusion, the increase in diastolic pressure yielded by esmolol therapy was clearly dose dependent (at all time points, p between low-dose and high-dose esmolol). During the postdrug period, hearts previously infused with the pacemaker current inhibitor had diastolic pressures that were significantly decreased compared with the corresponding baseline (predrug) values (p by paired t tests). The recovery of contractile function yielded slightly different patterns. At the end of drug infusion, the lowest values for dp/dt were seen, as expected, in high-dose esmolol-treated hearts (Table 1), whereas those of S were unchanged from baseline. In keeping with data on diastolic function, the recovery of dp/dt in esmolol-treated hearts was better in those receiving the lowest dose. After the discontinuation of drug infusion, function in the high-dose esmolol group gradually improved so that during the 1-hour postdrug period, the worst recovery of contractility became yielded by potassium arrest. At the opposite end, the best recoveries of function were seen in hearts infused with low-dose esmolol, high-dose nicorandil, and the pacemaker current inhibitor. In these three groups, postdrug values for dp/dt were not different from those of untreated controls. Esmolol and nicorandil yielded opposite dose-dependent effects on coronary flow (Table 1). Thus, within the esmolol group, the lower dose was associated with the highest coronary flows (p versus the higher dose both during and after drug infusion), whereas the nicorandil group featured an opposite pattern (ie, the higher dose was associated with significantly higher flows than the lower one [p and p during and after drug infusion, respectively]). In the S group, coronary flow decreased transiently during drug infusion and stabilized thereafter so that during the postdrug perfusion period, it was not different from that of control untreated hearts. Nor were these postdrug coronary flow values different from corresponding baseline data. Comment It is widely agreed that off-pump coronary artery bypass grafting operations are technically more challenging because of the greater difficulty in performing accurate distal anastomoses on a moving targeted coronary artery. This issue can be addressed by stabilizing devices using suction [9], pressure [10], or pharmacologic slowing of the heart rate. The present study was designed to character- Table 1. Comparative Effects of Drugs on Left Ventricular dp/dt and Coronary Flow Group (n 8) Baseline LV dp/dt (mm Hg/s) End of Drug Infusion Postdrug Period a Baseline Coronary Flow (ml/min) End of Drug Infusion Postdrug Period a Control 2, , , K arrest 2, , c Esmolol High 2, , b 1, d Low 2, , , f 51 1 g Nicorandil High 2, , , f 41 1 h Low 2, , , e S , , , a These values represent the average of values recorded over the 1-hour postdrug perfusion period. Values are given as mean standard error of the mean. b p versus all other groups. c p versus control, esmolol low, nicorandil high, and S ; p versus esmolol high; p versus nicorandil low. d p versus control; p versus nicorandil high; p versus S e p versus control; p versus nicorandil high; p versus S f p versus control, esmolol high, nicorandil low, and S g p versus all other groups. h p versus K ; p versus nicorandil low. LV dp/dt first derivative of left ventricular pressure.

4 Ann Thorac Surg BEL ET AL 1998;66: BRADYCARDIA DURING OFF-PUMP OPERATIONS 151 ize the class of drugs that could optimally combine achievement of bradycardia with preservation of function. The major finding is that the most effective approach for achieving this goal is based on pacemaker current inhibition. In this group, heart rate decreased by an average of 20% to 25% from baseline values. The clinical correlate (assuming the anesthesized patient s basal heart rate to be approximately 70 beats/min) would be a bradycardia in the range of 50 to 55 beats/min, which is a reasonable target rate for combining facilitated suturing with the lack of hemodynamic compromise. Until now, however, esmolol has probably been the most commonly used bradycardic drug during beating heart operations. This choice has been dictated by the cardioselectivity of this -blocker and the shortness of its half-life (a few minutes). In our clinical practice, however, the rate-lowering ability of esmolol has been found extremely variable, possibly because of the decreased -adrenoreceptor density in patients with a longstanding treatment by -blockers with intrinsic sympathicomimetic activity and in those with heart failure [11]. Unexpectedly, the present study failed to document a significant bradycardia in esmolol-treated hearts. This result is unlikely to be species dependent as the atria and, to a greater extent, the ventricles of the rabbit heart contain a high proportion of 1 -receptors [11], which are the elective targets for esmolol. Another explanation could be the use of a denervated heart but this hypothesis is challenged by the observation that the isolated heart has still a sympathetic tone supported by the noradrenaline release from the adrenergic nerve termini [12]. Indeed, in a similar isolated rabbit heart model, Ede and coworkers [4] have shown that esmolol can induce mechanical arrest (at a concentration of mol/l). A third explanation for the lack of bradycardia in esmolol-treated hearts is the inadequacy of the tested doses. The low dose was based on previously published data in a similar model [4]. The high dose, which was 20-fold greater (6 mmol/l), equally failed to slow the rate but was associated with markedly negative inotropic effects. It is likely that still higher doses, such as those recommended for intraoperative use during beating heart operations and which are in the molar range [1 3], would have induced a significant bradycardia. At these concentrations, however, esmolol could also exert nonspecific negative inotropic effects, which may be of concern in patients with an already tenuous hemodynamic status. Finally, because our protocol of drug administration had to be standardized, esmolol was given as a continuous infusion; we acknowledge that the efficacy of this mode of delivery might have been enhanced by an initial bolus administration. Potassium-channel openers have been proposed as alternates to potassium for inducing cardioplegia [13]. Their theoretical advantage is that these drugs hyperpolarize the cell membranes, which, in turn, shortens the duration of the action potential, reduces calcium influx, and limits the subsequent energy expenditure and tissue damage [13]. Our observation of low diastolic pressures and high contractile indices in nicorandil-treated hearts (in particular at the larger dose) is consistent with the ability of this drug to preserve function [5, 14], possibly as a result of reduced calcium overload [15]. Conversely, in spite of the previous findings that nicorandil shortens the time to arrest in a dose-dependent fashion [5, 6], in this study it failed to cause significant bradycardia, even at the highest dose of 1 mmol/l. It is noteworthy, however, that the cardioplegic properties of nicorandil have been demonstrated in hearts made globally ischemic after exposure to the drug, which was not the case in our experiments. Therefore, it is possible that under normoxic conditions such as those of the present protocol, still higher doses of nicorandil are required for slowing the heart rate. From a practical standpoint, however, such doses are completely irrelevant because of the nitratelike properties of nicorandil [16], which would cause profound hemodynamic effects before bradycardia can be observed. A different, and more physiologic, approach for inducing bradycardia is to reduce the heart sinus rate. One means of achieving this goal is to decrease the rate of diastolic depolarization that brings membrane potential from its most hyperpolarized level up to the threshold beyond which the action potential is fired and thus accounts for the pacemaking activity of sinoatrial node cells [7]. Four time-dependent currents have been implicated in this activity: the outward potassium current (I K ), two calcium currents (L-type and T-type), and the so-called inward pacemaker current (I f ). The latter, which is a mixed sodium potassium current activated by hyperpolarization, is considered the most important. In this study, we used S , a drug that has been shown to reduce the rate of spontaneous firing of action potentials primarily through a dose-dependent inhibition of I f [7]. At the selected concentration of 3 mol/l, the drug reduces I f by 50% [8], which is reflected by the significant bradycardia seen in the present experiments. The degree of this bradycardia is actually consistent with previous studies showing that the drug reduces heart rate by 20% to 30% of its baseline value [17 19]. For the sake of standardizing the protocol of drug administration, S was given as a continuous infusion over a 1-hour period and, consequently, it took approximately 30 minutes before bradycardia became apparent. During offpump operations, this time lag could be easily avoided by using a bolus-type of administration, which causes almost instantaneous bradycardia [8, 19]. The sustained decrease in heart rate seen with cessation of drug infusion is attributable to the prolonged duration of action of S This should not be a clinically relevant concern if the drug is to be used intraoperatively because of the reversibility of this bradycardia, if required, by temporary atrial pacing [17], provided the heart has been approached through a sternotomy. Interestingly, patch clamp experiments [8] have shown that S does not interact with the three other currents participating in the pacemaking activity. Thus, it has no effects on the calcium currents, which accounts for the absence of negative inotropic action, as demonstrated by our data and those reported by Simon and coworkers [17] in

5 152 BEL ET AL Ann Thorac Surg BRADYCARDIA DURING OFF-PUMP OPERATIONS 1998;66: resting and exercising conscious dogs. Likewise, it does not affect I K [8]. This is important because a prolongation of the action potential repolarization may increase the risk of ventricular arrhythmias. Other approaches can also be considered for slowing heart rate. Thus, the use of calcium-channel blockers has been advocated [20], but the negative inotropic effects of these drugs can raise concerns in patients with an already compromised left ventricular function. Adenosine is, at first look, another attractive candidate for inducing bradycardia through membrane hyperpolarization [21] and, in fact, Robinson and coworkers [22] have used boluses of adenosine for arresting the heart during the most critical part of anastomoses performed without pump support. A wide clinical acceptance of this approach remains, however, questionable in high-risk patients who may poorly tolerate repeated ventricular pauses and adenosine A 2 receptor-mediated vasodilation. A selective activation of the myocyte-bound adenosine A 1 receptors, otherwise shown to mediate the cardioprotective effects of ischemic preconditioning [23], would be more desirable but, unfortunately, none of these agonists is yet available for clinical use. There is no doubt that the results obtained in this isolated heart model should be extrapolated with caution to the clinical setting. However, the major advantage of this preparation is to provide carefully controlled experimental conditions that are well suited for the comparison of drugs. Indeed, as far as the pacemaker current inhibitor is concerned, it is noteworthy that the conclusions derived from isolated cardiac preparations have been fully supported by those drawn from in vivo dog [7] and pig [19] studies. This gives some support to our finding that the most effective approach for slowing heart rate without impairing function during operations on the beating heart seems to be the inhibition of the pacemaker current. The practical relevance of this finding stems from the fact that S is currently completing phase II clinical trials. Should it become soon available for human use, this drug might represent an effective component of the strategies designed to improve the technical accuracy of beating heart coronary artery bypass grafting procedures. Doctor Perrault is supported by the Clinician-Scientist program of the Medical Research Council of Canada. References 1. Sweeney MS, Frazier OH. Device-supported myocardial revascularization: safe help for sick hearts. Ann Thorac Surg 1992;54: Pirk J, Kellovsky P. An alternative to cardioplegia. Ann Thorac Surg 1995;60: Lönn U, Peterzen B, Granfeldt H, Casimir-Ahn H. Coronary artery operation with support of the Hemopump cardiac assist system. Ann Thorac Surg 1994;58: Ede M, Ye J, Gregorash L, et al. Beyond hyperkalemia: -blocker-induced cardiac arrest for normothermic cardiac operations. Ann Thorac Surg 1997;63: Qiu Y, Galinanes M, Hearse DJ. Protective effect of nicorandil as an additive to the solution for continuous warm cardioplegia. J Thorac Cardiovasc Surg 1995;110: Sugimoto S, Iwashiro K, Monti F, Dawodu AA, Sciariti M, Puddu PE. The risk of myocardial stunning is decreased concentration-dependently by K ATP channel activation with nicorandil before high K cardioplegia. Int J Cardiol 1995;48: Thollon C, Cambarrat C, Vian J, Prost JF, Peglion JL, Vilaine JP. Electrophysiological effects of S 16257, a novel sino-atrial node modulator, on rabbit and guinea-pig cardiac preparations: comparison with UL-FS 49. Br J Pharmacol 1994;112: Bois P, Bescond J, Renaudon B, Lenfant J. Mode of action of bradycardic agent, S 16257, on ionic currents of rabbit sinoatrial cells. Br J Pharmacol 1996;118: Jansen EWL, Gründeman PF, Borst C, et al. Less invasive off-pump CABG using a suction device for immobilization: the Octopus method. Eur J Cardiothorac Surg 1997;12: Boonstra P, Grandjean GJ, Mariani MA. Improved method for direct coronary grafting without CPB via anterolateral small thoracotomy. Ann Thorac Surg 1997;63: Brodde OJ. adrenoreceptors. In: Williams M, Glennon RA, Timmermans PBMWM, eds. Receptor pharmacology and function. New York: Marcel Dekker, 1989: Banerjee A, Locke-Winter C, Rogers KB, et al. Preconditioning against myocardial dysfunction after ischemia and reperfusion by an 1 -adrenergic mechanism. Circ Res 1993; 73: Cohen NM, Wise RM, Wechsler AS, Damiano RJ. Elective cardiac arrest with a hyperpolarizing adenosine triphosphate-sensitive potassium channel opener. A novel form of myocardial protection? J Thorac Cardiovasc Surg 1993;106: Mitani A, Kinoshita K, Fukamachi K, et al. Effects of glibenclamide and nicorandil on cardiac function during ischemia and reperfusion in isolated perfused rat hearts. Am J Physiol 1991;261:H Jovanovic A, Alekseev AE, Lopez JR, Shen WK, Terzic A. Adenosine prevents hyperkalemia-induced calcium loading in cardiac cells: relevance for cardioplegia. Ann Thorac Surg 1997;63: Grover GJ, Sleph PG, Parham CS. Nicorandil improves postischemic contractile function independently of direct myocardial effects. J Cardiovasc Pharmacol 1990;15: Simon L, Ghaleh B, Puybasset L, Giudicelli JF, Berdeaux A. Coronary and hemodynamic effects of S 16257, a new bradycardic agent, in resting and exercising conscious dogs. J Pharmacol Exp Ther 1995;275: Gardiner SM, Kemp PA, March JE, Bennett T. Acute and chronic cardiac and regional hemodynamic effects of the novel bradycardic agent, S 16257, in conscious rats. Br J Pharmacol 1995;115: Thollon C, Bidouard JP, Cambarrat C, et al. Stereospecific in vitro and in vivo effects of the new sinus node inhibitor ( )-S Eur J Pharmacol 1997;339: Nataf P, Lima L, Benarim S, et al. Video-assisted coronary bypass surgery: clinical results. Eur J Cardiothorac Surg 1997;11: De Jong JW, van der Meer P, van Loon H, Owen P, Opie LH. Adenosine as adjunct to potassium cardioplegia: effect on function, energy metabolism, and electrophysiology. J Thorac Cardiovasc Surg 1990;100: Robinson MC, Thielmeier KA, Hill BB. Transient ventricular asystole using adenosine during minimally invasive and open sternotomy coronary artery bypass grafting. Ann Thorac Surg 1997;63:S Downey JM, Thornton JD, Van Winckle DM, Stanley AWH, Olsson RA. Protection against infarction afforded by preconditioning is mediated by A 1 adenosine receptors in rabbit heart. Circulation 1990;84:350 6.