Ionic Basis of the Differential Effects of Intravenous Anesthetics on Erythromycin-induced Prolongation of Ventricular Repolarization in the Guinea Pig Heart 

Erythromycin (molecular weight, 733.9) was purchased from Fisher Scientific (Fair Lawn, NJ) and dissolved on the day of an experiment in the perfusion solution to a concentration of 100 micro Meter. Propofol (2,6-diisoprophylphenol; molecular weight, 178.3) was obtained commercially from Zeneca Pharmaceuticals (Wilmington, DE) as a sterile nonpyrogenic emulsion containing 10 mg/ml propofol. Intralipid 10%(Kabi Pharmacia, Clayton, NC), was obtained to exclude any effects attributable to propofol's vehicle. Ketamine (molecular weight, 274.2) was obtained commercially from Fort Dodge Laboratories (Fort Dodge, IA) as a 1:1 racemic mixture of 100 mg/ml ketamine. Sodium thiopental (molecular weight, 264.3) was purchased from Abbott Laboratories (North Chicago, IL) as a powder and mixed in sterile water to obtain a 25 mg/ml solution. Anesthetics were dissolved in the perfusion media to make 100 mM stock solutions and infused into the perfusion line to achieve the desired concentrations.

Isolation and Perfusion of Hearts. All protocols were reviewed and approved by the Animal Use Committee of the University of Florida Health Sciences Center. Hartley guinea pigs of either sex weighing 300–400 g were anesthetized with halothane (Halocarbon Laboratories, Rivers Edge, NJ) and killed by cervical dislocation. The hearts were rapidly removed and rinsed in ice-cold Krebs-Henseleit solution containing 117.9 mM NaCl, 4.8 mM KCI, 2.5 mM CaCl [center dot] 2H²O, 1.18 mM MgSO⁴[center dot] 7H²O, 1.2 mM KH²PO⁴, 0.5 mM Na²EDTA [center dot] 2H²O, 0.14 mM ascorbic acid, 5.5 mM glucose, 2 mM pyruvic acid (sodium salt), and 25 mM NaHCO sub 3. The ascending aorta was cannulated for perfusion of the coronary arteries at a constant flow of 8 ml/min with Krebs-Henseleit solution gassed continuously with 95% oxygen and 5% carbon dioxide. The oxygen tension, temperature, and pH of the Krebs-Henseleit solution were maintained at 500–600 mmHg, 36 +/- 0.5 [degree sign] Celsius, and 7.3–7.4, respectively. After dissection and instrumentation, the hearts were allowed to equilibrate for 30 min before experiments commenced.

Pacing and Electrophysiologic Measurements. Hearts were paced using an interval generator (A310 Accupulser, World Precision Instruments, Sarasota, FL) that delivered stimuli via a stimulus isolation unit (A360R, World Precision Instruments) as square-wave pulses lasting 3 ms and of twice the threshold intensity. The stimuli were delivered at a basic cycle length of 300 ms via a stainless steel, teflon-coated, bipolar electrode placed on the epicardium of the right ventricle.

Monophasic action potentials were recorded using pressure contact silver-silver chloride electrodes (Langendorff probe; EP Technologies, Sunnyvale, CA) placed on the epicardial surface of the left ventricle, as previously described. [13] The signals were amplified and filtered using an isolated biological amplifier (IsoDam, World Precision Instruments), and displayed in real time on a digital oscilloscope (model 2201; Tektronix Inc., Beaverton, OR). The amplitudes of the monophasic action potentials were determined from the diastolic baselines to the plateaus. Signals were considered adequate if they were stable for 10 min and their amplitudes exceeded 10 mV. Data were digitized using a DigiData 1200A digitizing system (Axon Instruments, Foster City, CA) and stored using the pClamp 6.1 data acquisition program (Axon Instruments) at 2 kHz for later analysis. Each data set was acquired as a group of 20 consecutive monophasic action potentials and subsequently averaged. The duration of the average monophasic action potential at 90% repolarization (MAPD⁹⁰) was measured using pClamp 6.1.

Experimental Protocols. After the control monophasic action potentials were recorded, the infusion of one of three randomly selected intravenous anesthetics (propofol, ketamine, or thiopental) was started. Each heart was treated with only one anesthetic. Anesthetics were infused for 20 min at 10, 25, and 50 micro Meter concentrations using a syringe infusion pump (sp 100i syringe pump; World Precision Instruments), and steady-state monophasic action potentials were recorded at each anesthetic concentration. After treatment of the hearts with all anesthetic concentrations, the anesthetic was discontinued and washed out for 60 min and monophasic action potentials recorded again. In a separate series of parallel experiments, drug-induced acquired LQTS was simulated by pretreating the hearts with erythromycin. After control monophasic action potentials were recorded, the hearts were perfused with Krebs-Henseleit solution containing erythromycin (100 micro Meter). Monophasic action potentials were recorded every 5 min, until steady-state effect was reached (Figure 1). Once the effects of erythromycin reached steady state (60 min), hearts were subjected to the same anesthetic drug administration protocol previously described.

Isolation of Ventricular Myocytes. Single ventricular myocytes were obtained from guinea pig hearts by enzymatic and mechanical dispersion, as previously described. [14] Briefly, the heart was quickly removed and perfused in a retrograde manner with oxygenated solution (100% oxygen at 36 +/- 0.5 [degree sign] Celsius) at a constant flow rate of 6 ml/min per gram of heart tissue. The perfusion solution contained 130 mM NaCl, 4.5 mM KCl, 3.5 mM MgCl², 0.4 mM NaH²PO⁴, 5 mM HEPES, 10 mM glucose, 20 mM taurine, 10 mM creatine, and 0.75 mM CaCl²; pH 7.25. After 5 min of perfusion with this solution, the perfusate was switched to a nominally Ca²+ free solution for 5 min. The hearts were then perfused for 10–20 min with the Ca²+ free solution (80 ml) containing 0.8 mg/ml collagenase (type 1; Worthington Biochemical Corp., Freehold, NJ) and 0.08 mg/ml protease (type XIV; Sigma Chemical Co., St. Louis, MO).

Thereafter, the heart was removed from the cannula. The ventricles were chopped coarsely with scissors, and placed into a beaker containing 5 ml Ca²+ free solution used for heart perfusion, the enzymes, and 6.4 mg/ml bovine serum albumin (fraction V, Sigma Chemical). The tissue was shaken in this solution for 3 min to disperse the cells mechanically. The cell suspension was filtered through a sterile gauze sponge and poured into 3.75 ml of a high-K sup +, low-Na sup 2+ solution containing 50 mM L-glutamic acid, 40 mM KCl, 10 mM HEPES, 0.5 mM EGTA, 20 mM taurine, 10 mM glucose, 3 mM MgCl², 70 mM KOH, 20 mM KH²PO⁴, and 6 mg/ml (100 mg/5 ml) bovine serum albumin; pH 7.2. The suspension was centrifuged for 2 min and the supernatant replaced with 2 ml of a high-K sup +, low-Na²+ solution and maintained at room temperature until needed.

Electrophysiologic Techniques. Aliquots of the cell suspension were transferred to a recording chamber mounted on the stage of an inverted microscope (Axiovert 10; Carl Zeiss, Thornwood, NY). A pipette connected to multiple temperature-controlled superfusion lines was positioned over the cell being studied to allow rapid (< 1 s) solution changes. The HEPES-buffered Tyrode's solution for superfusion of cells contained 130 mM NaCl, 5 mM KCl, 1.8 mM CaCl², 1 mM MgCl sub 2, 10 mM glucose, and 10 mM HEPES; pH 7.25. CdCl²(0.5 mM) was added to the external solution to eliminate the inward calcium current. Depending on the experimental protocol, solutions were modified by the appropriate addition (or substitution) of compounds, anesthetics, or both. To eliminate the known effects of nonphysiologic temperatures on current measurements, [15,16] the temperature of the superfusing solution was monitored using a digital thermometer (BAT-12; Physitemp Instruments, Clifton, NJ) and maintained at 36 +/- 0.5 [degree sign] Celsius.

The gigaseal technique for whole-cell patch-clamp recordings was used. [17] The patch microelectrodes were pulled from 1.5-mm KIMAX borosilicate capillary glass (Kimble-Kontes, Vineland, NJ) using a two-stage vertical puller (PP-83; Narishige USA, New York, NY). The patch microelectrodes had resistances of 3–5 M Omega when filled with the pipette-filling solution containing 107 mM potassium aspartate, 20 mM KCl, 1 mM MgCl², 5 mM HEPES, 5 mM EGTA, 1 mM CaCl², 4 mM Na sub 2 -ATP, and 0.4 mM GTP; pH 7.25. The voltage-clamp experiments were performed using an Axopatch ID amplifier (Axon Instruments). Data were monitored using an oscilloscope (5A26; Tektronix, Beaverton, OR), digitized online using a DigiData 1200A digitizing system (Axon Instruments), and stored on the hard drive of an IBM-compatible PC (P5–166; Gateway 2000, North Sioux City, ND). Voltage clamp protocols and off-line data analysis were performed using pClamp 6.1 software (Axon Instruments).

Recordings of I^Klwere obtained using a voltage ramp protocol wherein cells were held at -40 mV before their membrane potential was changed from -130 to 50 mV over 6 s in a linear manner. The I^Kwas studied in cells held at -40 mV (to inactivate I^Na) and depolarized by 600 ms pulses using 10 mV voltage steps to test potentials from -30 to 60 mV. The I^Kwas measured at the end of the 600 ms-long depolarizing pulses. All data were adjusted for a liquid junction potential of -10 mV.

All measurements are reported as means +/- SEM. Differences among multiple group means were made using univariate three-way repeated measures analysis of variance with one-way replication followed by Student-Newman-Keuls testing using SPSS version 7.5 statistical analysis software (SPSS Inc., Chicago, IL). Single mean comparisons were made using two-tailed, paired t testing. Differences between groups were considered significant at P < 0.05.

In hearts not treated with erythromycin, thiopental significantly prolonged MAPD⁹⁰in a concentration-dependent manner. The MAPD⁹⁰values after thiopental administration at 0, 10, 25, and 50 micro Meter were 163 +/- 6 ms, 180 +/- 7 ms, 193 +/- 7 ms, and 194 +/- 5 ms, respectively. The effect observed at 50 micro Meter thiopental was not significantly greater than that at 25 micro Meter. In contrast, propofol at lower concentrations (10 and 25 micro Meter) did not significantly affect MAPD⁹⁰. However, the highest concentration of propofol (50 micro Meter) significantly shortened MAPD⁹⁰(161 +/- 2 ms to 134 +/- 7 ms; P < 0.05). Ketamine had no significant effect. At all concentrations, differences between the effects of thiopental compared with those of propofol and ketamine on MAPD⁹⁰were statistically significant (Table 1). The MAPD⁹⁰changes were completely reversible for thiopental but only partially reversible for propofol. In four separate hearts, intralipid alone (propofol's vehicle) did not affect MAPD⁹⁰at rates corresponding to those of the highest concentration (50 micro Meter) of propofol (155 +/- 3 ms [control] vs. 158 +/- 2 ms [intralipid]; P = 0.22).

As shown in Figure 1, the effect of erythromycin on MAPD sub 90 reached steady state slowly. Erythromycin (100 micro Meter) prolonged MAPD⁹⁰in all hearts (n = 21, Figure 2). The MAPD⁹⁰increased from 155 +/- 2 ms to 171 +/- 2 ms during the 60-min erythromycin infusion (P < 0.001). No statistical difference in MAPD⁹⁰values before and after the 60-min erythromycin treatment, but before anesthetic administration, was noted among the three anesthetic groups.

In the continuous presence of erythromycin, thiopental further prolonged ventricular MAPD⁹⁰in a concentration-dependent manner (P < 0.05). The MAPD⁹⁰during treatment of the hearts with thiopental at 0, 10, 25, and 50 micro Meter was 168 +/- 4 ms, 191 +/- 5 ms, 210 +/- 6 ms, and 225 +/- 8 ms, respectively. Similar to hearts not treated with erythromycin, propofol (10 and 25 micro Meter) did not significantly affect MAPD90 in erythromycin-treated hearts, whereas the highest concentration of propofol (50 micro Meter) significantly shortened MAPD90 (171 +/- 4 ms to 163 +/- 5 ms, P < 0.05). Ketamine had no significant effect. At 25-micro Meter and 50-micro Meter concentrations, differences among the effects of thiopental, propofol, and ketamine on MAPD⁹⁰were statistically significant. The changes caused by erythromycin and the anesthetics were completely reversed during the 60-min washout period (Table 2). The MAPD90 for all drug groups after washout (156 +/- 4 ms) was not significantly different from the grand mean of control MAPD⁹⁰(155 +/- 2 ms), indicating complete reversal of drug effects on ventricular repolarization. Similarly, no differences in the washout values were observed among the anesthetic groups in hearts treated with erythromycin (Table 2).

After adjusting baseline values to account for the prolongation of MAPD⁹⁰caused by erythromycin, thiopental (25 and 50 micro Meter) caused greater lengthening of MAPD⁹⁰in the presence than in the absence of erythromycin (P = 0.018;Figure 3). In contrast, the shortening of MAPD⁹⁰caused by supraclinical concentrations of propofol (25 micro Meter and 50 micro Meter) were smaller in the presence than in the absence of erythromycin (P = 0.015;Figure 3).

(Figure 4) shows whole-cell voltage-clamp recordings of quasi-steady-state currents in guinea pig ventricular myocytes in response to a voltage ramp protocol. There was a large inward current shift at membrane potentials negative to -90 mV and a smaller outward current shift between -90 and -20 mV (primarily outward conduction through channels of the inward rectifier current I^KI. At more positive membrane potentials, there was a larger delayed outward current (I^K). Thiopental (50 micro Meter) significantly depressed inward I sub Kl at -120 mV and outward I^Klat -80 mV from -2,509 +/- 187 pA to -1,293 +/- 116 pA (53 +/- 4% of control) and from 500 +/- 44 pA to 108 +/- 29 pA (20 +/- 5% of control), respectively (n = 12). In addition, thiopental significantly inhibited the outward component of the ramp current at membrane potentials positive to -20 mV. For example, thiopental reduced potassium currents recorded at +40 mV (primarily representing I^K) from 901 +/- 120 pA to 524 +/- 73 pA (P < 0.05; n = 12). In contrast to isolated hearts, the effect of thiopental on the potassium currents in single ventricular myocytes was only partially reversible. As a typical example, the inward (-120 mV) and outward (-80 mV) I^Klcurrent on washout was -1966 +/- 265 pA (68 +/- 10% of control) and 140 +/- 76 pA (35 +/- 11% of control), respectively.

The effects of thiopental on I^Kwere further investigated by measuring the amplitude of I^Kat the end of 10-mV incremental, step depolarizations applied from a holding potential of -40 mV to potentials ranging from -30 to +60 mV (Figure 5). The results of the step protocol, a more sensitive method to assess time-dependent currents than the ramp protocol, confirmed the findings of the voltage ramp experiments. In seven cells, thiopental (50 micro Meter) changed the net I^Kin a similar manner as in the voltage ramp protocol. That is, thiopental (50 micro Meter) significantly decreased the amplitude of IK from 828 +/- 96 pA to 421 +/- 53 pA (53 +/- 3% of control) and from 154 +/- 31 pA to 60 +/- 13 pA (44 +/- 4% of control) at +50 and +10 mV, respectively (n = 7). Thus proportional inhibition of I^Kby thiopental tended to be greater at +10 mV than at +50 mV.

This study was designed specifically to investigate the effects of intravenous anesthetics in the setting of drug-induced, acquired LQTS. A major finding was that intravenous anesthetics caused different effects on ventricular repolarization in hearts treated with erythromycin, a drug known to cause QT prolongation and torsades de pointes by inhibiting the rapid component of the delayed rectifier potassium current (I^Kr). [11,18] Coincident with erythromycin administration, thiopental at clinically relevant, free concentrations (i.e., 7.2–60 micro Meter), [7,19,20]) unlike those of propofol or ketamine, further lengthened ventricular MAPD⁹⁰, in a concentration-dependent manner by a mechanism involving inhibition of not only the inward rectifier current (I^Kl, primarily between membrane potentials -90 to -20 mV) but also of the delayed rectifier potassium current (I^K, mainly at membrane potentials > -20 mV). The magnitude of the lengthening of MAPD⁹⁰caused by thiopental was significantly greater in the presence of erythromycin than in its absence, even after subtracting the lengthening caused by erythromycin alone.

Our findings are consistent with the results of other studies that have examined the effects of anesthetics on ventricular repolarization using isolated hearts or tissues. For example, thiopental (10–100 micro Meter) has been found to reversibly increase the action potential duration in isolated guinea pig perfused heart, [6] guinea pig [7] and canine [21] papillary muscle, and rabbit ventricular muscle. [22] In the present study, propofol shortened the action potential duration, but only at concentrations greater than the clinically relevant range of 1–10 micro Meter. [23–26] This finding is supported by previous studies in which propofol (<or= to 50 micro Meter) did not significantly change action potential duration. [7] On the other hand, the report of Baum [12] showing that propofol at 28 micro Meter blocks I^Kin guinea pig ventricular myocytes is difficult to explain in light of the results of previous studies [7] and the data presented here. We would expect an increase in MAPD⁹⁰if propofol blocked only IK. However, this paradox may be explained by the fact that anesthetics may simultaneously modulate the activity of several currents, and drug-induced changes in action potential duration reflect the net effect on several different individual currents. For example, propofol is also known to inhibit the cardiac L-type calcium channels (I^Ca,1)[27,28] In the present study, ketamine (<or= to 50 micro Meter) did not affect MAPD⁹⁰. This corresponds to the results of previous studies showing that ketamine affects action potential duration and diminishes I^Kl[12,29,30] and I^K[29,30] only at concentrations greater than the clinically relevant range of 3–90 micro Meter. [31,32]

Compared with the effects of anesthetics alone, erythromycin potentiated the prolongation of MAPD⁹⁰caused by thiopental (25 and 50 micro Meter) and attenuated the shortening caused by supraclinical concentrations of propofol (25 and 50 micro Meter), even after accounting for the prolongation caused by erythromycin treatment alone. The previous observation is in keeping with our finding that thiopental inhibits not only I^K, the mechanism by which erythromycin delays repolarization, but also depresses I^Klin isolated ventricular myocytes. Inhibition of both ionic currents may prolong MAPD⁹⁰synergistically.

In addition to these experimental studies, several clinical investigations have assessed the effects of thiopental and propofol on the QT interval duration. The results of these studies show that induction of anesthesia with either thiopental (5.0–7.5 mg/kg) or propofol (1.5–2.5 mg/kg) prolongs the corrected QT interval (QT^c) by approximately 10–36 ms. [4,5] The results for thiopental correspond well with our and others' experimental data, whereas the propofol-induced lengthening of the QT^cis not fully supported by data of the present and previous studies. [6,7] The differences between the clinical and experimental propofol data may be attributable to several factors. First, we simulated LQTS by treating the hearts with erythromycin, whereas the clinical investigators almost exclusively examined patients with normal repolarization. Of note, however, propofol (2 mg/kg) did not affect QTc in a subset of patients with prolonged QT sub c intervals. [4] Second, given the inaccuracy of Bazett's formula to adjust the QT interval for changes in heart rate, [33] the negative chronotropic effects of propofol, purportedly secondary to its direct actions (e.g., activation of M²-muscarinic cholinergic receptors [34] and inhibition of the L-type calcium current [27,28]) and indirect effects on autonomic nervous tone (e.g., gammaaminobutyric acid receptor-mediated anxiolysis leading to sympatholytic actions, [35] parasympathomimetic effects [36,37]) may have affected the clinical measurements of ventricular repolarization. Third, the measurement of the monophasic action potential duration used in the present study provides superior precision and accuracy for assessing changes in ventricular repolarization compared with the conventional measurements of the QT interval. [38]

In this study, we found that the thiopental-induced prolongation of ventricular repolarization is mediated, at least in part, by inhibition of outward potassium currents. Our results showed that both the inward rectifier I^Kland delayed rectifier IK were markedly suppressed by thiopental. Although the effect of thiopental on IKl has been reported, [8] ours is the first study to show that thiopental also affects IK at clinically relevant concentrations. This finding may have important clinical implications in the context of drug-induced LQTS. That is, by depressing both IKl and IK thiopental is likely to cause additional prolongation of the action potential duration, particularly at higher concentrations, even in the presence of a drug that produces only selective I^Klor I^Kblock. On the other hand, many previous studies have shown that thiopental inhibits the L-type calcium channel. [27,28] Antagonism of the L-type calcium channel effectively terminates early afterdepolarizations, the putative electrophysiologic substrate of torsades de pointes. [39–41] This finding, combined with preliminary results from our laboratory showing that thiopental prolongs ventricular repolarization in a frequency-independent manner, [6] may reduce the risk for the development of torsades de pointes related to thiopental-induced action potential prolongation. Frequency-independence refers to a phenomenon whereby drug-induced lengthening of the action potential duration does not depend on the underlying heart rate. Usually drugs that prolong the action potential by selectively blocking a potassium channel (e.g., IK block by d-sotalol) exhibit greater effects on ventricular repolarization at slow heart rates (i.e., reverse frequency-dependence behavior). [13,42] The lack of effect of heart rate on thiopental-induced lengthening of action potential duration [6] is in keeping with the findings of previous studies that more electrophysiologically complex drugs such as amiodarone and ATI-2001 exhibit frequency-independent behavior on ventricular repolarization. [13]

Although our results corresponded mainly with those of Pancrazio et al., [8] there were some differences. First, we found that the outward component of I^Klwas more sensitive to thiopental than the inward component was. They showed that thiopental (30 micro Meter) caused a slightly greater inhibition of the outward component of I^Kl(66% of control) compared with the inward component of I^Kl(54% of control). Second, the onset of the effect of thiopental on I^Klwas faster in our experiments and reached maximum changes in 2–3 min, whereas previous investigators pretreated cells for 20 min with thiopental. Third, we observed that thiopental inhibited not only I^Klbut also I^K. These differences can be explained by different experimental conditions and drug concentrations. Most importantly, our experiments were performed at 36 [degree sign] Celsius, whereas other researchers did their experiments at 22 [degree sign] Celsius. [8,15,16]

When the results of our study are interpreted, some potential limitations should be noted. First, the results from guinea pig heart may not be directly extrapolated to humans. For example, the guinea pig heart does not possess the transient outward potassium current (I^to), which is partly responsible for ventricular repolarization in humans. [43] Second, any supposition that thiopental may be more dysrhythmogenic than propofol or ketamine in hearts with delayed ventricular repolarization must be tempered by the fact that effects of anesthetics on innervated (in vivo) hearts may differ from those on denervated (in vitro) Langendorff hearts. For example, although we found that clinically relevant concentrations of propofol did not affect MAPD sub 90 in isolated hearts paced at a constant rate, the phenol-derivative may delay repolarization in vivo by slowing heart rate not only via several direct mechanisms previously described [27,28,34] but also by modulating autonomic tone. [35–37] On the other hand, by using an isolated heart model and controlling heart rate, we could directly assess the effects of anesthetics on ventricular repolarization without the confounding influences of autonomic nervous tone on heart rate and ventricular action potential duration. Third, we chose to simulate acquired LQTS in the guinea pig heart by treating the hearts with erythromycin. Whether these results are directly applicable to delayed ventricular repolarization caused by different agents such as drug-induced I^Klblock (e.g., terfenadine)[44] is not known. Fourth, the effects of anesthetics on action potentials in patients with congenital LQTS (vs. acquired LQTS) remains to be established. Several genetic mutations cause distinct subtypes of congenital LQTS, including SCN5A (persistent inward sodium current), KVLQT1, and HERG (inhibition of the potassium currents). [45] The acquired and congenital forms of LQTS vary dramatically with respect to aggravating factors and treatment. For example, enhanced sympathetic innervation to the heart may precipitate torsades de pointes in patients with congenital LQTS. [45] In contrast, isoproterenol is used to treat acquired LQTS. Therefore, anesthetics known to increase sympathetic drive to the heart should be used cautiously in patients with congenital LQTS.

Prolongation of ventricular repolarization (QT interval) is a well known risk factor for the development of malignant ventricular dysrhythmias and sudden death. [39,45] Anesthetics that further prolong ventricular repolarization in patients with acquired LQTS may predispose patients to the development of torsades de pointes-type ventricular tachydysrhythmias during the perioperative period. Therefore, the selection of an anesthetic induction agent may be an important consideration with regard to dysrhythmogenesis and clinical outcome. Although our data suggest that thiopental should be used cautiously in the setting of acquired (drug-induced) LQTS, additional correlative clinical studies are warranted before definitive recommendations can be made.