Abstract
Recently, there has been considerable attention focused on drugs that prolong the QT interval of the electrocardiogram, with the H1-receptor antagonist class of drugs figuring prominently. Albeit rare, incidences of QT prolongation and ventricular arrhythmias, in particular torsade de pointes, have been reported with the antihistamines astemizole and terfenadine and more recently with loratadine. The most likely mechanism for these drug-related arrhythmias is blockage of one or more ion channels involved in cardiac repolarization. Several studies have demonstrated block of multiple cardiac K+ channels by terfenadine, includingIto, Isus,IK1, and IKr or human ether-a-go-go-related gene (HERG). In contrast to terfenadine, previous studies have shown the antihistamine loratadine to be virtually free of cardiac ion channel-blocking effects. This disparity in the lack of any significant cardiac ion channel-blocking effect and the existence of numerous adverse cardiac event reports for loratadine prompted the comparison of the human cardiac K+channel-blocking profile for loratadine and terfenadine under physiological conditions [37°C, holding potential (Vhold) = −75 mV] with the whole-cell patch-clamp method. Isolated human atrial myocytes were used to examine drug effects on Ito,Isus, and IK1, whereas HERG was studied in stably transfected HEK cells. In contrast to previous studies in nonhuman systems and/or under nonphysiological conditions, terfenadine (1 μM) had no effect onIto, Isus, orIK1 at pacing rates up to 3 Hz. Similar results were found for 1 μM loratadine. However, both drugs potently blocked HERG current amplitude, with a mean IC50 of 173 nM for loratadine and 204 nM for terfenadine (pacing rate, 0.1 Hz). Neither drug exhibited any significant use-dependent blockage of HERG (pacing rates = 0.1–3 Hz). These results point to a similarity in the human cardiac K+ channel-blocking effects of loratadine and terfenadine and provide a possible mechanism for the arrhythmias associated with the use of either drug.
Recently, there has been considerable attention focused on drugs that prolong the QT interval of the electrocardiogram. Examples can be found in many therapeutic classes, the danger being that excessive QT prolongation, which reflects delayed myocardial repolarization, may lead to potentially lethal ventricular arrhythmias (Stratmann et al., 1987; Zipes, 1987; Zehender et al., 1991; Benedict, 1993). The likelihood of developing deleterious adverse cardiac effects when exposed to QT-prolonging drugs can be enhanced under certain conditions such as electrolyte abnormalities, metabolic disturbances, and preexisting medical conditions such as heart disease and congenital long QT syndrome (Jackman et al., 1988).
The H1-receptor antagonist class of drugs has figured prominently in regulatory agency concerns regarding QT-prolonging drugs. Several cases of QT prolongation and ventricular arrhythmias, in particular torsade de pointes, have been reported with the antihistamines astemizole and terfenadine (Craft, 1985; Bishop and Gaudry, 1989; Davies et al., 1989; Monahan et al., 1990; Lindquist and Edwards, 1997). These reports prompted warning of potentially serious adverse cardiac events with these agents. The most likely mechanism for these drug-related arrhythmias is blockage of one or more ion channels involved in cardiac repolarization. Several studies have demonstrated blockade of multiple cardiac K+ channels by terfenadine, including Ito,Isus,IK1, andIKr or human ether-a-go-go-related gene (HERG) (Table 1). In contrast to terfenadine, studies with nonhuman systems and/or under nonphysiological conditions have shown the antihistamine loratadine to be virtually free of cardiac ion channel-blocking effects (Table2). However, recent case reports have appeared describing both atrial and ventricular arrhythmias associated with loratadine usage (Good et al., 1994; Haria et al., 1994; Lindquist and Edwards, 1997; de Abajo et al., 1999). Furthermore, a search of the World Health Organization database reveals an adverse cardiac event profile for loratadine, which includes ventricular arrhythmias, similar to that of terfenadine (Lindquist and Edwards, 1997), although examples of torsade de pointes associated with loratadine use have not been reported. This disparity in the lack of any significant cardiac ion channel-blocking effect and the existence of numerous adverse cardiac event reports for loratadine prompted the comparison of loratadine and terfenadine under physiological conditions with human cardiac ion channels. The effects of these drugs on ion channels were examined in either isolated human cardiac myocytes or a human cell line expressing the human K+ channel, HERG. This is the first study to examine and compare the ion channel-blocking profile and the rate dependence of loratadine and terfenadine under physiological conditions and with human cardiac ion channels.
Materials and Methods
Human tissue was obtained, in accordance with Tulane University School of Medicine institutional guidelines. Myocytes were isolated from specimens of human right atrial appendage obtained during surgery from hearts of five patients (ages 43–69 years) undergoing cardiopulmonary bypass. All atrial specimens were described as grossly normal at the time of excision and had no evidence of dilation on ECG, and all patients had normal P waves on ECG. Despite these qualifications, the atrial specimens may not represent “normal” atrial tissue. Some patients had received cardioactive drugs including Ca2+ channel blockers and digitalis. The cell-isolation procedure has been described in detail (Crumb et al., 1995a).
Transfection and Cell Culture.
HEK 293 cells stably expressing HERG mRNA were maintained in minimum essential medium with Earle's salts supplemented with nonessential amino acids, sodium pyruvate, penicillin, streptomycin, and fetal bovine serum.
Solutions.
Isolated human atrial myocytes or HEK cells were superfused with an “external” solution that consisted of 137 mM NaCl, 4 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 11 mM glucose, 10 mM HEPES; adjusted to a pH of 7.4 with NaOH. Glass pipettes were filled with an “internal” solution that consisted of 120 mM K-aspartate, 20 mM KCl, 4 mM Na-ATP, 5 mM EGTA, 5 mM HEPES; adjusted to a pH of 7.2 with KOH. Loratadine and terfenadine were kindly provided by Almirall Prodesfarma (Barcelona, Spain) and dissolved in 100% dimethyl sulfoxide (DMSO) to make concentrated stock solutions (1 mM). Loratadine was added to the bath solution from this concentrated stock (final DMSO concentration less than 0.1%) or from another more diluted stock solution (100 μM) made in distilled deionized water.
Conditions.
Experiments were performed in the presence of 200 μM Cd2+ to block the l-type Ca2+ current and at a holding potential of −75 mV. All experiments were performed at 36 ± 1°C. The transient outward current was measured by subtracting the amplitude of the current measured at the end of a depolarizing voltage pulse (sustained current) from peak current amplitude. The sustained current was measured at the end of a depolarizing pulse to +60 mV. Outward HERG tail current was measured to assess drug effects.
Acceptable atrial myocytes were rod shaped and lacked any visible blebs on the surface. Currents were measured with the whole-cell variant of the patch-clamp method (Hamill et al., 1981). Currents were digitized at 3 to 5 kHz and filtered at 1 kHz. Pipette tip resistances were approximately 1.0 to 2.0 MΩ when the pipettes were filled with the internal solution. Analog capacity compensation and 40 to 60% series resistance (Rs) compensation was used in all experiments to yield voltage drops across uncompensated Rs of less than 3 mV. Paired and unpaired Student's t tests were used for statistical analysis. Data are presented as means ± S.E.
Results
Figure 1 illustrates the ion channel-blocking effects of loratadine and terfenadine at various pacing rates on Ito,Isus, andIK1 recorded from isolated human atrial myocytes. These experiments were performed under physiological conditions of temperature (36 ± 1°C), holding potential (−75 mV), and external [K+] (4 mM). As indicated, even at pacing rates as high as 3 Hz, neither 1 μM terfenadine nor 1 μM loratadine had any effect on these human K+currents. In contrast, both agents markedly inhibited HERG current amplitude recorded from stably expressing HEK cells (Fig.2). HERG current was measured as an outward tail current elicited on repolarization to −40 mV from an immediately preceding depolarizing voltage pulse to +10 mV. At a pacing rate of 0.1 Hz, there was a significant reduction in outward HERG tail current amplitude compared with control on addition of either 100 nM loratadine (48.9 ± 3.8%, n = 9) or 100 nM terfenadine (41.1 ± 5.1%, n = 6) (p < .01). Fits of mean dose-response curves (n = 5–10) with the equation
The results reported in this study are very different from those ofTaglialatela et al. (1998), who recently reported that loratadine at a concentration of 3 μM produced only an approximately 30% reduction in HERG current amplitude in a human neuroblastoma cell line expressing HERG. The conditions used in this study (37°C, 400-ms depolarizing pulses, 4 mM [K+]o, outward tail currents) and those in the Taglialatela et al. study (22°C, 10-s depolarizing pulses, 100 mM [K+]o, inward tail currents) were quite different. When these conditions were mimicked in this study, results similar to those in the Taglialatela et al. study were observed, with 3 μM loratadine producing a 30.2 ± 2.1% (n = 4) reduction in HERG current amplitude (Fig.3). These results were dramatically different from those obtained under more physiological conditions in this study, where 100 nM loratadine produced a greater than 45% reduction in HERG current amplitude, and 3 μM loratadine is predicted to produce an approximately 80% reduction (Figs. 2C and 3).
Discussion
The results of this study indicate that, of the four human cardiac K+ currents tested under physiological conditions, only HERG was blocked by either terfenadine or loratadine. The lack of any observed block of Ito,Isus, andIK1 by 1 μM terfenadine (Fig. 1) is in contrast to previous reports that indicate a 10 to 40% reduction in these currents with the same concentration of terfenadine (Table 1). Although it is not clear, these differences may be the result of species differences in cardiac ion channels or in experimental conditions (Table 1). It is interesting that terfenadine had no blocking effect on Isus in human atrial cells because it has been shown to block Kv1.5, a channel cloned from human heart and expressed in mammalian cell lines (Rampe et al., 1993). The lack of terfenadine blockade ofIsus may reflect the fact that Kv1.5 is not the only current in human atria and may not be the major current underlying Isus (Crumb et al., 1995b). In human ventricle, a Kv1.5-like current cannot be recorded (Mays et al., 1995). These observations make it unlikely thatIsus,Ito, orIK1 are targets for terfenadine and loratadine at concentrations associated with arrhythmias. The lack ofIto blockade by terfenadine reported in this study is in contrast to a recent publication by Crumb (1999) in which terfenadine was found to potently and rate-dependently blockIto in human atrium recorded at 22°C. The likely reason for the disparity discussed inResults is that this study examined the effects of terfenadine on Ito at 37°C, at which drug unbinding kinetics are faster than those recorded at 22°C.
Another striking result of this study is the similarity in the HERG-blocking potency and rate dependence exhibited by loratadine and terfenadine (Fig. 2). Whereas the results indicating an IC50 of 204 nM for terfenadine blockade of HERG are consistent with other studies on terfenadine (Table 1), previous studies have failed to observe any HERG or nativeIKr blocking action associated with submicromolar concentrations of loratadine (Table 2). This disparity might be explained by differences in experimental conditions. Indeed, when the experimental conditions of a previous study suggesting little HERG-blocking activity by loratadine were mimicked, similar results were obtained (Fig. 3). In their study, Taglialatela et al. (1998)performed experiments at room temperature, currents were elicited by very long depolarizing pulses (10 s), a high external [K+] was used (100 mM), and inward tail currents elicited by hyperpolarizing pulses to −140 mV were used to measure drug effects on current amplitude. In contrast, in this study, experiments were performed at 37°C, currents were elicited by much shorter voltage pulses (400 ms), 4 mM external [K+] was used, and outward tail currents elicited by pulses to −40 mV were used to measure drug effects on current amplitude. It is not known which of the experimental differences may be involved in the observed differences in loratadine-blocking affinity of HERG, but it has been reported that elevating external [K+] decreases drug affinity forIKr (Yang et al., 1996). Alternatively, the difference discussed in Results may be due to species differences in IKr. It has been shown that subtle changes in the amino acid sequence of ERG (ether-a-go-go-related gene) can result in dramatic changes in ERG pharmacology, as evidenced by the human and bovine forms of this channel, where a single amino acid change can result in a 100-fold difference in the sensitivity to theIKr blocker dofetilide (Ficker et al., 1998).
Loratadine blockage of HERG described herein provides a mechanism for the arrhythmias reported in association with loratadine usage (Good et al., 1994; Haria et al., 1994; Lindquist and Edwards, 1997; de Abajo et al., 1999). Indeed, a study examining the World Health Organization database for reporting of adverse drug reactions suggests that the incidence of arrhythmias reported in association with loratadine use is similar to that of terfenadine (Lindquist and Edwards, 1997), although life-threatening examples of torsade de pointes and incidences of sudden death have not been described with loratadine usage, whereas they have been with terfenadine use. Furthermore, a recent study including nearly 200,000 patients indicates that the risk of ventricular arrhythmias associated with terfenadine usage was no different than that observed for loratadine (de Abajo et al., 1999). The demonstration that terfenadine and loratadine share similar HERG-blocking potencies and rate dependence, over a concentration range associated with arrhythmias (Hilbert et al., 1987, 1988; Davies et al., 1989), is consistent with these clinical observations. Further studies correlating HERG blockade with myocardial concentrations of terfenadine and loratadine associated with arrhythmias would be beneficial. Differences in the incidence of severe, life-threatening arrhythmias between loratadine and terfenadine may rely more on differences in attainable myocardial concentrations of drug than differences in HERG blockade.
Acknowledgments
The author thanks Drs. Craig January and Zhengfeng Zhou for kindly supplying HERG transfected HEK cells, and Drs. Nabil Munfakh, Herman A. Heck, and Lynn H. Harrison for kindly providing atrial specimens.
Footnotes
-
Send reprint requests to: William J. Crumb, Jr., Ph.D., Department of Pediatrics, Division of Cardiology, Tulane University School of Medicine, 1430 Tulane Ave., New Orleans, LA 70112-2699. E-mail: wcrumb{at}tmcpop.tmc.tulane.edu
- Abbreviations:
- HERG
- human ether-a-go-go-related gene
- DMSO
- dimethyl sulfoxide
- ERG
- ether-a-go-go-related gene
- Received May 17, 1999.
- Accepted August 31, 1999.
- The American Society for Pharmacology and Experimental Therapeutics