Prevention of Torsade de Pointes during the Pharmacologic Treatment

Mitamura H

Prevention of TdP during AF therapy

Review Article

Prevention of Torsade de Pointes during the Pharmacologic Treatment of Atrial Fibrillation Hideo Mitamura MD, PhD Department of Cardiology, Saiseikai Central Hospital

Potassium channel blockers are sometimes effective in the rhythm management of persistent atrial fibrillation. However, on rare occasions, a life-threatening episode of torsade de pointes develops during treatment with these agents. Among several precautions against this tragedy, preventing exaggerated QT prolongation is of paramount importance. One of the difficulties in monitoring the QT interval during atrial fibrillation is the fact that its length varies depending not only on the fluctuating preceding intervals, but on other factors such as neurohumoral activation. In fact, even at the same preceding RR interval, the QT interval may vary from day to night, between atrial fibrillation and sinus rhythm. Accordingly, the QTc interval may not be a reliable parameter for predicting proarrhythmic events. As the result, close monitoring of the QT interval itself, just like INR monitoring during warfarin treatment, is mandatory both before and during treatment with these agents. (J Arrhythmia 2010; 26: 5–15)

Key words: Torsade de pointes, Atrial fibrillation, QT interval, Antiarrhythmic drug

Sunny and dark sides of pharmacologic rhythm management of atrial fibrillation Atrial fibrillation (AF) remains a challenging arrhythmia despite significant progress being made in suppression of this arrhythmia both pharmacologically and nonpharmacologically. Acute AF is often successfully suppressed with various Na channel blockers,1) and if it is repetitive and symptomatic, pulmonary vein isolation with radiofrequency ablation is becoming a powerful option in eliminating this arrhythmia. In contrast, when AF persists longer and electrical remodeling develops, pharmacological rate control of this arrhythmia is sometimes preferred.1,2) One of the reasons for favoring the conservative approach to persistent AF is that, as

atrial electrical remodeling progresses, the anti-AF efficacy of Na channel blockers wanes,3,4) making rhythm management more difficult. On the other hand, some of the K channel blockers such as amiodarone, sotalol, or dofetilide have been reported to be effective in terminating such persistent AF, although their conversion rates were unimpressive at less than 30%.5–7) Notably, 1 to 2 months of daily administration of bepridil, an IKr and other multichannel blocker, has demonstrated surprisingly high efficacies of 58 to 69% in terminating persistent AF,8,9) and its clinical use has recently been approved in Japan for the indication of pharmacologic rhythm management of persistent AF. This recommendation, however, is not straightforward, and needs to accompany a warning to those

Address for correspondence: Hideo Mitamura MD, PhD, Director, Cardiac Research Center, Department of Cardiology, Saiseikai Central Hospital, 1-4-17, Mita, Minato-ku, Tokyo 108-0073 Japan. E-mail: [email protected]

5

J Arrhythmia

Vol 26 No 1 2010

who plan to prescribe such K channel blockers for this purpose. One should be aware that most of these agents exert their action on myocardial cells of both the atrium and the ventricle. While the effect of prolongation of the atrial refractory period by the drug may be antiarrhythmic and beneficial, the same or even exaggerated effect on the ventricle can be proarrhythmic and unwanted. The result of the latter is torsade de pointes (TdP), a notorious life-threatening arrhythmia with a characteristic electrocardiographic manifestation associated with prolonged ventricular repolarizaton. This proarrhythmic effect of K channel blockers has been seen not only in the treatment of ventricular tachyarrhythmias, but also in the treatment of supraventricular tachyarrhythmias,10,11) a more unexpected and less desirable result. For example, in a multicenter, double-blind, placebo-controlled, dose-ranging study of dofetilide called SAFIRE-D, 3 proarrhythmic events leading to ventricular fibrillation including 1 fatal case occurred in 241 patients receiving different doses of dofetilide.7) In another multicenter, double-blind, placebo-controlled, dose-ranging study in Japan called J-BAF which comprised 90 patients with persistent atrial fibrillation, 1 patient died of proarrhythmic effect of bepridil within three months of follow-up.12) The treatment should not be worse than the disease itself. However, the prevention of TdP in this particular setting of atrial fibrillation is sometimes complicated and difficult as will be discussed subsequently. Precipitating factors for TdP TdP is caused by abnormal automaticity and subsequent reentrant arrhythmias in an electrophysiological substrate of markedly prolonged and dispersed repolarization of myocardial cells. The initiating automatic activity, or triggered activity, is the result of early afterdepolarization, which characteristically develops during the phase 3 of the action potential only when its duration is abnormally prolonged13,14) (Figure 1). The prolongation of the action potential can often be brought by inhibiting outward currents. Therefore, those agents blocking outward K currents, which theoretically prolong the refractory period of myocardial cells and is expected to prevent reenty, are at the same time a typical culprit responsible for this proarrhythmic phenomenon. We have long intended to avoid this tragic complication during the treatment with K channel blockers in various ways. The blockade of K currents with these agents is apparently dependent

6

0

100 mV

200 msec Figure 1 Quinidine-induced (Adopted from 13)

early

afterdepolarization

Action potentials were recorded from a single canine Purkinje cell at cycle lengths from 1,000 to 8,000 msec. Note the action potential duration increased as the cycle length was increased to the critical point when early afterdepolarization developed during phase 3.

on its concentration. Therefore, it is a matter of course to limit the dose of these agents after considering the body weight and age as well as liver and renal functions of the patient. It may be better to withhold the administration of these agents to those in whom the baseline duration of the action potential is prolonged due to congenitally defective genes regulating the action potential (long QT syndrome) or adjust the dose in similar patients with acquired defective genes encoding the K channel, such as those developing heart failure.15) Caution should be exercised when prescribing these agents to female patients who are known to be more susceptible to TdP than male patients, a difference most likely related to gender-specific hormones.16) In addition, since the action potential duration increases as the serum potassium becomes lower, the proarrhythmic adverse events with K channel blockers occur more frequently in the presence of hypopotassemia as is often the case during treatment with diuretics.17) It has been suggested that the serum potassium level should always be kept above 4.0 mEq/L. Drug-to-drug interaction is always a matter of concern. It is not unusual for a patient to receive a new drug for noncardiac diseases from another doctor. It will be helpful to check these risk factors for TdP systematically both before and after the treatment with K channel blockers has been instituted. These factors can easily be memorized by alphabet acronyms as listed in Table 1.

Mitamura H

Prevention of TdP during AF therapy

Risk factors for TdP

Table 1 A B C D E F G H

aged bradycardia cardiomyopathy drugs combined excessive dose female genetics hypopotassemia

QT interval as the key determinant of TdP Even when all these precautions and recommendation are taken into consideration by each prescribing doctor, the prediction and prevention of TdP remain difficult. As the response of the myocardial cells to K channel blockers is quite variable and sometimes unpredictable, all we need is to monitor the patient carefully, sometimes continuously, sometimes repeatedly. No doubt the most important parameter to monitor is the QT interval on the surface electrocardiogram, which reflects summation of the action potential duration of the ventricular myocardial cells. It is at the same time, a marker for the degree of K channel blockade. If one is treating supraventricular arrhythmias with K channel blockers, QT prolonagation is not a requisite since the target for prolongation of action potential duration is not the ventricular cells but the atrial cells whose action potential duration is neither apparent nor reflected on the surface electrocardiogram. On the other hand, QT prolongation is not only a marker for antiarrhythmic effect of K channel blockers against ventricular arrhythmias, but more importantly a marker for the risk of developing TdP in any clinical settings. In a case-control study comparing 30 cases of bradyarrhythmias complicated by TdP with 113 cases of uncomplicated bradyarrhythmias, cutoff value for the QT interval that distinguished the former from the latter was 570 msec with 90.0% sensitivity and 86.7% specificity.18) In another single center survey of 459 patients receiving bepridil for the treatment of AF or flutter, Yasuda et al have reported that TdP developed in 4 patients during an average follow up of 20 months.19) In addition to the fact that 3 of them had structural heart disease, all 4 had hypopotassemia with serum K levels 3.8 mEq/L or lower. QT intervals before developing TdP were consistently 0.56 sec or longer.

Monitoring the QT interval is not just recommended but is mandatory each time a K channel blocker is prescribed for the treatment of arrhythmias, whether supraventricular or ventricular. However, monitoring the QT interval during AF is, unlike that during sinus rhythm, much more challenging. It is true that it is sometimes difficult to precisely measure the QT interval in the presence of fine waves of AF. The automatic measurement of QT intervals using a computer lacks reliability particularly during AF. However, the problem with AF is far more complicated. Variable QT intervals during AF Even in a single subject and at a constant serum potassium level, the action potential duration and the QT interval vary substantially within a day. Characteristically, the QT interval prolongs when the heart rate is slow as during the night, and shortens as the heart rate becomes faster as during activities in the daytime. This phenomenon is seen even in patients without K channel blockers. Notably, however, when a K channel blocker is administered, this pattern of reverse rate dependency (the faster the rate, the less K channel blockade) gets magnified,20) so that a marked prolongation of QT interval becomes apparent sometimes during the night (Figure 2). This pattern of reverse rate dependency has been reported to be more significant in the presence of structural heart disease21) (Figure 3). It follows that the risk of developing early afterdepolarization and TdP becomes higher during bradycardia18,22) particularly in patients with structural heart disease. In a previously cited series of Yasuda et al, heart rates were 48 or less in all the 4 patients developing TdP.19) Although bradycardia is typically seen during the night, a long pause triggering TdP can also develop in other situations. A long pause generated after premature contractions sometimes gives rise to a similar but transient milieu for developing TdP. During AF, there are wide fluctuations of the RR interval, sometimes accompanying oscillatory shortlong-short cycle length patterns. These patterns are known to set the stage for TdP even during the daytime.23) The extreme of this condition is brought when AF or atrial tachyarrhythmias convert into sinus rhythm. After a prolonged sinus node suppression induced by rapid atria excitation, a substantially long pause follows once the tachyarrhythmia terminates, which may predispose the patient to TdP (Figure 4). This phenomenon could even be accentuated in the presence of sick sinus syndrome, which at times is induced by antiarrhythmic drugs.

7

J Arrhythmia

15:37

Vol 26 No 1 2010

HR 71/min

QT 0.42 sec

HR 33/min

QT 0.72 sec

CH1 CH2 CH3 3:17 CH1 CH2

Figure 2 QT intervals during day (top) and night (bottom) in a patient receiving bepridil

CH3

A significant QT prolongation was revealed when the heart rate was decreased or SA block developed during night.

0.60

HR: 100

80

60

50 beat/min

QT Interval (sec)

0.55 0.50

0.45 pffiffiffiffiffiffiffi Figure 3 QT- RR relationships based on measurements from 24-hour Holter electrocardiograms (Adopted from 21)

0.40 0.35 0.30 0.7

0.8

0.9 Preceding

1.0 R-R

Since the longer the QT interval, the higher the likelihood of developing TdP, finding the longest QT interval after an intensive search for a single complex preceded by the longest RR interval is extremely important. It would be easy to understand that the measurement of QT intervals on a routine electrocardiogram of a short duration recorded in the daytime in front of a doctor or technician wearing a white lab coat may not be sufficient and may be misleading. Obtaining a longer trace by using a Holter monitoring is thus preferred for this purpose even though there is a practical limitation as to how often it should be recorded. With this technique, however, it has been demonstrated that the develop-

8

1.1

1.2

The slope was larger in patients with structural heart disease receiving bepridil (closed triangle, 0.57, n ¼ 14) than those in patients without structural heart disease (open triangle, 0.34, n ¼ 15) or control subjects not receiving bepridil (open circle, 0.32, n ¼ 29) (p < 0:01).  p < 0:01 vs control subjects,  p < 0:01 vs patients without structural heart disease

ment of TdP was closely linked to the longest RR intervals (Figure 5). More complex behavior of QT intervals It would be interesting to note that the RR interval during treatment with K channel blockers is not a single determinant of the QT interval and thus the development of TdP. In a detailed analysis of the relationship between the QT and RR intervals preceding the onset of ventricular arrhythmias in a patient with acquired prolongation of ventricular repolarization, Gilmour et al have found that the QT intervals were variable even if the immediately

Mitamura H

Prevention of TdP during AF therapy

I II III aVR aVL aVF V1 V2 Figure 4 Pause-dependent TdP

initiation

of

V4 V5 V6

QT interval (msec)

Note a marked QT prolongation in a sinus beat following a pause after termination of atrial tachyarrhythmia. This patient was relatively stable on chronic bepridil treatment until when tryptanol was prescribed by a dermatologist for painful herpes zoster infecton which precipitated this event.

V3

600 550 500 450 400 350 300 2

4

6

8

10 12 Time (hours)

14

16

18

20

2

4

6

8

10 12 Time (hours)

14

16

18

20

Figure 5 QT (top) and RR (bottom) intervals during ambulatory ECG monitoring in a patient with drug-induced QT prolongation (Adopted from 24)

RR interval (msec)

2000 1800 1600 1400 1200 1000 800

Episodes of ventricular arrhythmias including TdP are indicated by the vertical lines.

preceding RR intervals were similar24) (Figure 6). They also demonstrated that the slope of the linear relation between the RR and QT intervals was less steep for the 10 intervals that preceded an episode of ventricular arrhythmias than for earlier intervals (Figure 7). In other words, the QT interval became longest at any given RR interval just before the onset of an episode of ventricular arrhythmias. The QT

interval was indeed a function not only of the immediately preceding RR interval, but also of previous RR intervals. In fact, it was dependent on the cumulative effect of the bradycardia that precedes an episode of ventricular arrhythmias and on the tachycardia that occurs during the episode. It is true that the TQ interval, the interval from the end of T wave to the biginning of QRS complex

9

J Arrhythmia

A

Vol 26 No 1 2010

C

580

1 2 3 4

QT Interval (msec)

560 540

5 6

520 500 480 460 440 420

20 40 60 80 100 120 Number of complexes

B

0

200

400 600 Time (ms)

800

D 1500

1450

1

RR Interval (msec)

1400

Figure 6 QT (A) and RR (B) intervals preceding the onset of ventricular arrhythmias in the same patient as in Figure 5. (Adopted from 24)

2

1350

3

1300 1250 1200

4

1150

5 6

1100 1050 1000

20 40 60 80 100 120 Number of complexes

400

500 600 Time (ms)

reflecting the excitable period of the ventricle, rather than RR interval that affects the QT interval of the subsequent beat. Thus, even if the preceding RR intervals are the same, the QT interval of the subsequent beat may differ if QT intervals of the preceding beats vary. Moreover, alteration of autonomic nervous system tone in response to changes in heart rate may also contribute to modulation of the QT interval. It has been demonstrated that sympathetic stimulation will activate IKs and shorten the action potential duration, potentially counteracting the effect of K channel blockers.25) If such tone can change during AF, the RR and QT relationship may also be variable. Furthermore, the profile of QT-RR hysteresis, i.e., the speed with which the QT interval adapts to heart rate changes, is considered to be highly individual with intrasubject stability and intersubject variability.26) To make matters more complicated, this relationship between the QT and the preceding RR intervals varies significantly once AF is converted into sinus

10

700

The ECG recordings for the six circled complexes in A and B are shown in C, aligned by the R wave. The terminal portions of the T waves for these complexes are shown in D at an expanded time scale. The asterisks indicate the fiducial point on the T wave for the QT measurements. The QT interval increased progressively, resulting in a premature ventricular complex (recording 6). Note that RR intervals 2 and 3 were similar, as were RR intervals 4 and 5, yet the corresponding QT intervals differed.

rhythm (Figure 8). An important finding of exaggerated QT prolongation after cardioversion of AF has been reported by Choy et al27) In their study, dofetilide has been shown to cause only minor QT prolongation during AF, but significantly more QT prolongation when given to the same patients after cardioversion to sinus rhythm. Their data indicates that the QT response to dofetilide in AF does not predict the response after restoration to sinus rhythm, and that exaggerated QT response may occur in predisposed individuals. Furthermore, Darbar et al have also demonstrated that cardioversion of AF acutely increases the QT interval and the steepness of the QT-RR slope which was seen in subjects taking sotalol or quinidine but not amiodarone or flecainide28) (Figure 9). Interestingly, in accordance with these observations, it has been demonstrated that persistent AF is associated with reduced risk of TdP in patients with drug-induced long QT syndrome.29) Based on these observations, one should accept that the RR and QT relationship is not fixed but

Mitamura H

Prevention of TdP during AF therapy

.60

QTe vs RRe

.58

QTl vs RRl

Figure 7 Relation between the QT and RR intervals for the periods of sinus rhythm separating the seven episodes of ventricular arrhythmias shown in Figure 6 (Adopted from 24)

QT Interval (sec)

.56

The QT/RR relations are shown for the last 10 intervals preceding episodes of ventricular arrhythmias (late [QTl and RRl]) and for the remaining intervals (early [QTe and RRe]). There was an upward shift of the QT interval for a given RR interval preceding the onset of ventricular arrhythmias. The upward shift of the QT interval was greater for short RR intervals than for long RR intervals.

.54 .52 .50 .48 .46 .44 1

1.1

1.2

1.3

1.4

1.5

1.6

RR Interval (sec)

During AF

After conversion to SR

V1

V2

Figure 8 A marked QT prolongation (0.68 sec) was exposed after AF was converted to sinus rhythm (SR)

V3

rather dynamic. Accordingly, the corrected QT interval or QTc interval which is considered to reflect this relationship, is also a vulnerable parameter. Although the QTc interval has long been advocated as a guide to predict development of TdP,30) it needs to be reconsidered, particularly in the setting of AF. Characteristically, the QTc interval during AF is not only difficult to measure, but even if it is measured properly and calculated by a variety of formulas, it can still vary within a short time period and thus is an inappropriate, sometimes misleading parameter in predicting the risk of TdP in patients receiving K channel blockers.31) Furtheore, both the Bazett and Fridericia correction formulas are reported to overestimate the change in QT interval at heart rates from 60 to 120 bpm or higher32) while

underestimating it at heart rates below 60 bpm. A recent AHA/ACCF statement indicated that the QTc interval longer than 500 msec is a proarrhythmic marker.30) If the Bazett formula is inappropriately applied, a QT duration of 612 msec could become the longest acceptable interval for a beat preceded by the RR interval of 1.5 sec during AF, which is apparently misleading and even dangerous. The application of these formulas to AF or sinus rhythm with frequent ectopic beats as well as extreme bradycardia or tachycardia should be avoided or should require extreme caution. It should be remembered that early depolarization develops at the critical action potential duration (Figure 1), or at the critical QT interval on the surface electrocardiogram (Figure 6), and never develops at the critical ‘‘cor-

11

J Arrhythmia

Vol 26 No 1 2010

0.6

0.5

Bazett Fridericia

0.4

QT (sec)

0.3

0.2 300

600

900

1200

1500

0.6

0.5

Bazett Fridericia

0.4

0.3

0.2 300

600

900

1200

1500

RR (msec) Figure 9 QT-RR interval relationship in AF (top) and in sinus rhythm (bottom) for subjects taking sotalol or quinidine (circle) and those taking amiodarone or flecainide (square) (Adopted from 28)

Predicted lines according to the Bazett (QTc=QT/RR^ (1/2)) or Friedricia (QTc=QT/RR^ (1/3)) formulae are also shown for reference.

rected QT interval’’. The QTc interval, as a matter of fact, is just a marker for abnormal rate adaptation of the QT interval, which can be diagnostic of genetic long QT syndrome. This corrected interval would be more appropriately used for the purpose of screening the patient in sinus rhythm for such hereditary abnormalities susceptible to K channel blockers. Monitoring QT intervals, practical considerations As repeatedly stressed, the QT interval rather than the QTc interval should be monitored during the treatment with K channel blockers, particularly in the presence of AF. For treatment of AF with K channel blockers, the longest QT interval, not the average QT interval should be sought in an electrocardiogram since the longest but not the average QT

12

interval is the one most closely linked to the development of TdP. The technical approaches to measuring QT intervals or analyzing the QT morphology have been discussed elsewhere and are not the subject of this review.18,33–35) In routine clinical practice, QT intervals are measured visually and manually, which can be approximated to a 0.04 sec level of accuracy. Unfortunately, there is no established threshold below which prolongation of the QT interval is considered free of proarrhythmic risk, and no clearcut consensus on the degree of drug-induced QT prolongation that should require drug discontinuation. Based on previously reported data, this author would like to suggest the following guidelines, which can be applied to patients with or without AF. Whenever the longest QT interval exceeds 0.52 sec, the prescribing doctor should be alerted that

Mitamura H

TdP is impending. Once the interval exceeds 0.56 sec, the dose should either be withheld or decreased to bring back the interval below this level. Importantly, when the QT interval of any single beat exceeds 0.60 sec, the risk of developing TdP is so high that not only should the culprit drug be discontinued, but the patient is advised to be hospitalized and placed under continuous electrocardiographic monitoring. The management toward prevention of TdP in a particular patient is continuous and laborious. Both before and during the treatment with K channel blockers, one is advised to confirm whether any of the risk factors discussed earlier with alphabet acronyms (Table 1) exist that can modify the effect of the drugs. It may be helpful to measure the QTc interval before prescribing a K channel blocker only if there is sinus rhythm. If it is abnormally prolonged from the beginning, the patient may have a reduced repolarization reserve and genetic predisposition to drug-induced TdP,36) and the administration of such a drug should be discouraged. Once the drug has been started, one should always keep in mind that the QT interval is not fixed, but varies constantly second by second, hour by hour, day by day, and year by year. The initial loading or titration period of drug administration requires particular caution. In one study comprising 120 patients with atrial arrhythmias, Chung et al have demonstrated that new or increased ventricular arrhythmias developed in 7 patients at 2:9  1:8 days after initiation of sotalol.37) They recommended hospital admission and at least 72 hours of monitoring for initiation of sotalol. In other drugs such as bepridil, it takes longer before the plateau of the QT prolongation is reached, thus frequent visits to an outpatient clinic or use of transtelephonic monitoring instead of in-hospital monitoring would be acceptable. Even after the peak of QT prolongation has been reached, it is recommended to take occasional 24 hour Holter monitoring, particularly for patients with AF. The QT interval should be measured whenever any events occur that modify the action potential duration. For example, when diarrhea occurs and hypopotassemia develops, it can be expected that the QT interval should be prolonged further even in the presence of the same dose of K channel blockers. Or when a patient visits another doctor for another disease to receive other drugs with variable degrees of K channel blocking action, the QT interval should be monitored intensively. As described earlier, when persistent AF has been converted to sinus rhythm, one should measure the QT interval as soon as possible.

Prevention of TdP during AF therapy

AF sometimes facilitates the development or progression of heart failure, while the latter can increase the risk for the development of AF. With or without AF, however, if a patient is taking K channel blockers, it happens that TdP occasionally develops after aggressive treatment of heart failure. Hypopotassemia is often induced by diuretics. The heart rate usually slows down during the recovery phase of heart failure, which may unmask bradycardia-dependent QT prolongation. The increased neurohumoral compensatory mechanism at the peak of acute heart failure may be inactivated during convalescence, which could also unmask background QT prolongation. On top of these vulnerable parameters, diseased ventricular myocardial cells suffering left ventricular failure characteristically have been known to exhibit prolonged action potential durations.14) All these factors in concert prolong QT interval significantly, predisposing to TdP when the patient is about to get out of the nightmare of dyspnea. The approach to coping with this moving target is analogous to the control of INR during warfarin administration for patients with AF. It takes days to weeks to achieve the therapeutic plateau. Therefore, frequent measurements of the QT interval are suggested particularly during the initial titration period. As INR has to be adjusted to a level that is not too high to cause bleeding and not too low to cause cerebral embolism, the QT interval has to be adjusted somewhere between the extremes. As INR changes frequently despite keeping the same dose of warfarin, the QT interval also fluctuates easily despite taking the same dose of a K channel blocker. Accordingly, the QT interval should be measured routinely, as with INR, every time AF patients taking these agents visit the clinic. Both INR and the QT interval are vulnerable to drug-to-drug interaction, so that caution should be advised against taking new drugs without confirming the safety of the combination, particularly when those drugs are to be prescribed by independent doctors. Just like INR, repeated measurements of the QT interval are required every time a new drug has been added. And just as INR is measured at the time of a bleeding event, the measurement of the QT interval is a must when a patient taking K channel blockers develops syncope or presyncope. Interestingly, the administration of warfarin is recommended both before and after conversion of persistent AF is attempted. Thus it is not uncommon for both K channel blockers and warfarin to be prescribed together to the same patient. It would be helpful in such a situation to remember an old saying

13

J Arrhythmia

Vol 26 No 1 2010

that all drugs are poisons with some beneficial side effects. Every doctor prescribing drugs to patients should be aware of this proverb, which is particularly true during the treatment of AF with K channel blockers and warfarin.

References 1) Ogawa S, Yamashita T, Yamazaki T, et al: Optimal treatment strategy for patients with paroxysmal atrial fibrillation. J-RHYTHM study. Circ J 2009; 73: 242–248 2) The AFFIRM Investigators: A Comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002; 347: 1825–1833 3) Reisinger J, Gatterer E, Heinze H, et al: Prospective comparison of flecainide versus sotalol for immediate cardioversion of atrial fibrillation. Am J Cardiol 1998; 81: 1450–1454 4) Sato T, Mitamura H, Kurita Y, et al: Electropharmacologic effects of pilsicainide, a pure sodium channel blocker, on the remodeled atrium subjected to chronic rapid pacing. J Cardiovasc Pharmacol 2001; 38: 812–820 5) Hohnloser SH, Kuck KH, Lilienthal J, et al: Rhythm or rate control in atrial fibrillation: Pharmacological Intervention in Atrial Fibrillation (PIAF): a randomized trial. Lancet 2000; 356: 1789–1794 6) Singh BN, Singh SN, Reda DJ, et al for the Sotalol Amiodarone Atrial Fibrillation Efficcy Trial (SAFE-T) Investigators: Amiodarone versus sotalol for atrial fibrillation. N Engl J Med 2005; 352: 1861–1872 7) Singh BN, Zoble RG, Yellen L, et al: Efficacy and safety of oral dofetilide in converting to and maintaining sinus rhythm in patients with chronic atrial fibrillation or atrial flutter. The symptomatic atrial fibrillation investigative research on dofetilide (SAFIRE-D) study. Circulation 2000; 102: 2385–2390 8) Fujiki A, Tsuneda T, Sugao M, et al: Usefulness and safety of bepridil in converting persistent atrial fibrillation to sinus rhythm. Am J Cardiol 2003; 92: 472–475 9) Nakazato Y, Yasuda M, Sasaki A, et al: Conversion and maintenance of sinus rhythm by bepridil in patients with persistent atrial fibrillation. Circ J 2005; 69: 44–48 10) Falk RH: Proarrhythmia in patients treated for atrial fibrillation or flutter. Ann Intern Med 1992; 117: 141– 150 11) Hohnloser SH, van de Loo A, Baedeker F: Efficacy and proarrhythmic hazards of pharmacologic cardioversion of atrial fibrillation: prospective comparison of sotalol versus quinidine. J Am Coll Cardiol 1995; 26: 852–858 12) Yamashita T, Ogawa S, Sato T, et al for the J-BAF investigators: Dose-response effects of bepridil in patients with persistent atrial fibrillation monitored with transtelephonic electrocardiograms. A multicenter, randomized, placebo-controlled, double-blind study (JBAF Study). Circ J 2009; 73: 1020–1027 13) Roden DM, Hoffman RF: Action potential prolongation and induction of abnormal automaticity by low quinidine concentrations in canine Purkinje fibers. Relationship to potassium and cycle length. Circ Res 1985; 56: 857–867

14

14) Roden DM: Cellular basis of drug-induced torsades de pointes. Brit J Pharmacol 2008; 154: 1502–1507 15) Tomaselli GF, Marban E: Electrophysiological remodeling in hypertrophy and heart failure. Cardiovasc Res 1999; 42: 270–283 16) Makkar RR, Fromm BS, Steinman RT, et al: Female gender as a risk factor for torsades de pointes associated with cardiovascular drugs. JAMA 1993; 270: 2590–2597 17) Yang T, Roden DM: Extracellular potassium modulation of drug block of IKr. Implications for torsade de pointes and reverse use-dependence. Circulation 1996; 93: 407– 411 18) Topilski I, Rogowski O, Rosso R, et al: The morphology of the QT interval predicts torsade de pointes during acquired bradyarrhythmias. J Am Coll Cardiol 2007; 49: 320–328 19) Yasuda M, Nakazato Y, Sasaki A, et al: Clinical evaluation of adverse effects during bepridil administration for atrial fibrillation and flutter. Circ J 2006; 70: 662–666 20) Okada Y, Ogawa S, Sadanaga T, et al: Assessment of reverse use-depedent blocking actions of class III antiarrhythmic drugs by 24-hour Holter electrocardiography. J Am Coll Cardiol 1996; 27: 84–89 21) Sadanaga T, Ogawa S: Bepridil produces prominent bradycardia-dependent QT prolongation in patients with structural heart disease. J Electrocrdiol 2007; 40: 426– 431 22) Davidenko JM, Cohen L, Goodrow R, et al: Quinidineinduced action potential prolongation, early afterdepolarizations, and triggered activity in canine Purkinje fibers. Effects of stimulation rate, potassium, and magnesium. Circulation 1989; 79: 674–686 23) Locati EH, Maison-Blanche P, Dejode P, et al: Spontaneous sequences of onset of torsade de pointes in patients with acquired prolonged repolarization: quantitative analysis of Holter recordings. J Am Coll Cardiol 1995; 25: 1564–1575 24) Gilmour Jr RF, Riccio ML, Locati EH, et al: Time- and rate-dependent alterations of the QT interval precede the onset of torsade de pointes in patients with acquired QT prolongation. J Am Coll Cardiol 1997; 30: 209–217 25) Vanoli E, Priori SG, Nakagawa H, et al: Sympathetic activation, ventricular repolarization and IKr blockade: implications for the antifibrillatory efficacy of potassium channel blocking agents. J Am Coll Cardiol 1995; 25: 1609–1614 26) Malik M, Hnatkova K, Novotny T, et al: Subject-specific profiles of QT/RR hysteresis. Am J Physiol Heart Circ Physiol 2008; 295: H2356–H2363 27) Choy AM, Darbar D, Dell’Orto S, et al: Exaggerated QT prolongation after cardioversion of atrial fibrillation. J Am Coll Cardiol 1999; 34: 396–401 28) Darbar D, Hardin B, Harris P, et al: A rate-independent method of assessing QT-RR slope following conversion of atrial fibrillation. J Cardiovasc Electrophysiol 2007; 18: 636–641 29) Darbar D, Kimbrough J, Jawaid A, et al: Persistent atrial fibrillation is associated with reduced risk of torsades de pointes in patients with drug-induced long QT syndrome. J Am Coll Cardiol 2008; 51: 836–842

Mitamura H

30) Drew BJ, Ackerman MJ, Funk M, et al: Prevention of torsade de pointes in hospital settings: a scientific statement form the American Heart Association and the American College of Cardiology Foundation. Circulation 2010; 121: epub 31) Rawles JM: The QT interval in atrial fibrillation. Br Heart J 1989; 61: 510–513 32) Indik JH, Pearson EC, Fried K, et al: Bazett and Fridericia QT correction formulas interfere with measurement of drug-induced changes in QT interval. Heart Rhythm 2006; 3: 1003–1007 33) Hondeghem LM: Thorough QT/QTc not so thorough: removes torsadogenic predictors from the T-wave, incriminates safe drugs, and misses profibrillatory drugs. J Cardiovasc Electrophysiol 2006; 17: 337–340 34) Kirchhof P, Franz MR, Bardai A, et al: Giant T-U waves precede torsades de pointes in long QT syndrome. A

Prevention of TdP during AF therapy

systematic electrocardiographic analysis in patients with acquired and congenital QT prolongation. J Am Coll Cardiol 2009; 54: 143–149 35) Katoh T: Clinical background and evaluation of druginduced prolongation of QT interval. J Arrhythmia 2009; 25: 56–62 36) Itoh H, Sakaguchi T, Ding WG, et al: Latent genetic backgrounds and molecular pathogenesis in drug-induced long QT syndrome. Circ Arrhythm Electrophysiol 2009; 2: 511–523 37) Chung MK, Schweikert RA, Wilkoff BL, et al: Is hospital admission for initiation of antiarrhythmic therapy with sotalol for atrial arrhythmias required? Yield of in-hospital monitoring and prediction of risk for significant arrhythmia complications. J Am Coll Cardiol 1998; 32: 169–176

15