- Email: [email protected]
High-dose bolus lidocaine for chemical cardioversion of atrial fibrillation: A prospective, randomized, double-blind crossover trial
High-dose bolus lidocaine for chemical cardioversion of atrial fibrillation: A prospective, randomized, double-blind crossover trial Nassir F. Marrouche, MD, Ramakota K. Reddy, MD, Ann K. Wittkowsky, PharmD, and Gust H. Bardy, MD Seattle, Wash
Background Most drugs used for chemical cardioversion of atrial fibrillation have significant proarrhythmia risk and require close monitoring after administration. Lidocaine has few of the proarrhythmic concerns of most antiarrhythmic drugs and, at high bolus doses, prolongs the atrial refractory period well enough to be effective in converting atrial fibrillation to sinus rhythm. This finding has been previously demonstrated in a dog model. We sought to confirm the animal findings in human beings with lidocaine doses of 1.5 to 2.5 mg/kg.
Methods Twenty patients with atrial fibrillation scheduled for elective cardioversion were enrolled in this study. In a randomized, double-blind, crossover study design, each patient received intravenous bolus lidocaine or saline. Patients were observed for 10 minutes after the initial bolus to assess efficacy. The second test drug was then delivered if the first was unsuccessful at cardioversion.
Results All 20 patients received both lidocaine and saline placebo therapy in a crossover manner. None of the 20 patients converted to sinus rhythm with either therapy. The 95% confidence interval for effectiveness of lidocaine in this population was 0% to 14%.
Conclusion In this population of patients referred for elective cardioversion of atrial fibrillation, high-dose bolus lidocaine was ineffective in converting patients to sinus rhythm. Although this study was not sufficiently powered to rule out a low efficacy of lidocaine (<15%) or a higher efficacy in certain subgroups of atrial fibrillation, routine use of lidocaine for this indication is not warranted. (Am Heart J 2000;139:e6.)
Atrial fibrillation is the most common arrhythmia in the United States, affecting 0.4% of the adult population. It accounts for 3.5% of all cardiology hospital admissions and 35% of admissions for arrhythmia. At least 18% of first strokes can be attributed to atrial fibrillation.1,2 To decrease stroke risk, improve hemodynamics, and restore normal sinus node activity, cardioversion of atrial fibrillation is part of common clinical practice.3 Cardioversion is usually accomplished by use of transthoracic shocks with appropriate anesthesia or by pharmacologic means with appropriate follow-up monitoring for proarrhythmia. The availability of an inexpensive agent for conversion of atrial fibrillation without sedation or prolonged observation would allow cardioversion to be attempted in an outpatient clinic setting rather than as a hospital procedure. From the Department of Medicine, University of Washington. Supported by a fellowship award from the American Heart Association, Washington affiliate. Submitted August 9, 1999; accepted January 24, 2000. Reprint requests: Gust H. Bardy, MD, Clinical Coordinating Center, 7900 East Greenlake Dr North, Suite 300, Seattle, WA 98103-4819. E-mail: [email protected] Copyright © 2000 by Mosby, Inc. 0002-8703/2000/$12.00 + 0 4/90/106169 doi:10.1067/mhj.2000.106169
Intravenous lidocaine is a readily available agent familiar to all physicians. It is generally used for ventricular arrhythmias because of its minimal effect on atrial myocardium at low to moderate doses. At higher doses, achievable with rapid bolus administration, lidocaine has the property of prolonging the atrial refractory period, similar to class III agents.12,13,15,16 Such high serum levels cannot be maintained with continuous infusion because of neurologic side effects, but cardioversion should require only transient prolongation of the atrial refractory period, similar to the effect of electrical cardioversion. In fact, in a dog model of atrial fibrillation, bolus lidocaine was effective in terminating all of 101 episodes of atrial fibrillation.4 Consequently, we sought to evaluate the efficacy of high dose intravenous bolus lidocaine for cardioversion of atrial fibrillation in human beings in a prospective, placebo-controlled manner.
Methods Patient population Patients referred for elective electrical or chemical cardioversion of atrial fibrillation were screened for the study. Institutional review board approval was obtained for the study,
Table I. Patient demographics General No. of patients Male Age (y) Height (cm) Weight (kg) Cardiovascular history Coronary artery disease Nonischemic dilated cardiomyopathy Valvular disease Without structural heart disease (lone atrial fibrillation) Noncardiovascular history Hypertension Diabetes mellitus Hypothyroidism Hyperthyroidism Smoker Alcohol use (>1 drink/d) Left atrium size (mm) Atrial fibrillation duration (d) Left ventricular ejection fraction (%) NYHA classification I II III
20 15 62 ± 12 174 ± 4 90 ± 21 8 (40%) 3 (15%) 2 (10%) 7 (35%) 10 (50%) 2 (10%) 2 (10%) 1 (5%) 3 (15%) 6 (30%) 51 ± 6 187 ± 317 47 ± 17 10 (50%) 8 (40%) 2 (10%)
Values are n or n (%).
and all patients had to provide informed written consent for participation. Patients with atrial flutter at the time of cardioversion were excluded. Patients with a history of seizures, liver disease, or sensitivity to lidocaine were also excluded. Of 40 patients approached for the study, 2 were excluded because they spontaneously converted to normal sinus rhythm before randomization, 3 were excluded because they had atrial flutter immediately before the procedure, 3 were excluded because of a history of seizure disorder, and 1 was excluded because of a history of liver disease. Of the remaining 31 patients, 11 declined to participate in the study; thus 20 patients were enrolled.
Lidocaine administration protocol Half of the 20 patients were assigned a lidocaine dose of 1.5 mg/kg and half a dose of 2.5 mg/kg. At the time of cardioversion, syringes with equal volume of lidocaine and saline placebo were prepared and coded in a manner not revealing their contents by the hospital research pharmacy. In a random and blinded manner, an initial treatment with either saline placebo or lidocaine was injected over a 10second period with continuous electrocardiographic and noninvasive blood pressure monitoring for 10 minutes. If patients did not convert to normal sinus rhythm within 10 minutes, they were crossed over to the other arm and the second 10 mL bolus of either saline or lidocaine was infused over a 10-second period. Patients were monitored for an additional 10-minute period. If, after crossover, the patient remained in atrial fibrillation, a transthoracic electrical cardioversion was performed. Continuous single-lead electrocardiographic recordings
were made from 1 minute before administration of each agent to 5 minutes after administration of the last test drug. The primary end point was conversion of atrial fibrillation to normal sinus rhythm within 5 minutes of drug administration. Vital signs were recorded at baseline and at 2 minutes after drug administration to assess the hemodynamic consequence of bolus lidocaine.
Statistical methods The study design was based on the premise that an efficacy of 20% would be required for the acceptance of lidocaine as a useful agent for conversion of atrial fibrillation, assuming the side effects from the bolus were tolerable. An initial sample size of 50 was chosen to provide 90% power to identify a 20% efficacy above placebo at significance of .05. However, after the initial 20 patients were enrolled with no successful conversions in either arm, an interim analysis revealed a 99% certainty that the efficacy would be <20% and a 95% certainty that the efficacy would be <15%. The study was therefore terminated for lack of efficacy.
Results Clinical characteristics Of the 20 patients enrolled, 15 (75%) were men. The mean age was 62 ± 12 years (range 46 to 84 years), the mean height was 174 ± 9 cm (range 157 to 193 cm), and the mean weight was 90 ± 21 kg (range 52 to 150 kg). Ten patients (50%) had a history of hypertension, whereas only 2 patients (10%) had diabetes. Two patients (10%) had hypothyroidism and were receiving replacement therapy, and 1 patient (5%) had a history of treated hyperthyroidism. Only 3 patients (15%) were smokers. One patient (5%) admitted to heavy alcohol use (>3 drinks per day), whereas 5 patients (25%) were moderate drinkers (1 to 2 drinks per day), and 14 patients (70%) drank rarely or not at all. The duration of atrial fibrillation was 187 ± 317 days with a range of 0 to 1140 days. Two patients had postoperative atrial fibrillation. Sixteen patients (75%) were anticoagulated at the time of cardioversion with mean international normalized ratio of 2.77 ± 0.55. All the patients who were anticoagulated had therapeutic international normalized ratios of ≥2.0 for 3 weeks before cardioversion. The 4 patients who were not anticoagulated had atrial fibrillation of <48-hour duration. For 14 patients (70%), this was the first cardioversion; 6 patients (30%) had 1 or more previous successful electrical cardioversions (mean 1.8). Underlying structural heart disease included coronary artery disease in 8 patients (40%), nonischemic dilated cardiomyopathy in 3 (15%), and isolated valvular disease in 2 (10%). The remaining 7 patients (35%) had lone atrial fibrillation without other identifiable underlying cardiac disease. Left ventricular function was normal in 11 patients (55%) and only mildly depressed, with ejection fraction of 40% to 50% in 3 patients (15%). Six patients (30%) had ejection fraction <40%. Functional
class was normal (New York Heart Association [NYHA] class I) in 10 patients (50%); 8 patients (40%) were NYHA class II, and 2 patients (10%) were NYHA class III. No patient had class IV symptoms. Echocardiographic data were available in 18 of the 20 patients. The mean left ventricular end-diastolic dimension was 56 ± 8 mm, and the mean left ventricular endsystolic dimension was 38 ± 11 mm. The mean fractional shortening was 33% ± 12%. There was left atrial enlargement of >40 mm in all but 1 patient, with mean left atrial size of 51 ± 6 mm. Patients demographics are listed in Table I.
Adverse effects All patients in the protocol reported neurologic symptoms from the lidocaine bolus; 3 patients also reported symptoms from placebo. Fifteen patients (75%) noted dysarthria after the lidocaine bolus, 11 patients (55%) noted paresthesia, 3 (15%) had confusion or anxiety, 2 (10%) had shortness of breath or dry mouth each, and 1 patient (5%) had chest pain. Two patients specifically noted that time seemed compressed, feeling only 20 to 30 seconds had elapsed at 6 minutes. Nine patients (45%) noted visual or ocular changes or hallucinations, with one patient, a pediatric ophthalmologist, specifically describing “vertical oscillopsia.” Central nervous system depression and nausea each occurred in 5 of the 10 patients in the high-dose (2.5 mg/kg) group; neither of these was noted in patients in the lowdose (1.5 mg/kg) group. Six of the 10 patients in the low-dose group noted euphoria; this was not noted in any of the patients in the high-dose group. All the symptoms from the lidocaine bolus resolved within 10 minutes. Three patients reported effects during the placebo infusion, 1 with paresthesia, 1 with shortness of breath, and 1 with nausea. The patient with paresthesia from placebo received placebo first in the crossover design, whereas the other 2 patients received lidocaine first.
Cardioversion results In all 20 patients, both lidocaine and placebo failed to convert atrial fibrillation to normal sinus rhythm, even transiently. Transthoracic cardioversion was subsequently performed in 19 of the 20 patients, with 1 patient refusing to proceed to DC cardioversion because of persistent nausea. Sinus rhythm was restored in 13 patients (68%) with a mean of 1.5 shocks. The other 6 patients (32%) remained in atrial fibrillation.
Hemodynamic effects No significant hemodynamic changes were registered after the administration of 1.5 mg/kg and 2.5 mg/kg lidocaine bolus. Documented heart rate was 86.5 ± 14.8 beats/min at baseline and 91.2 ± 10.9 beats/min 2 minutes after administration of 1.5 mg/kg lidocaine bolus (P = .4). The systolic and diastolic blood pressure changed
nonsignificantly from 130.6 ± 22.3 to 136.4 ± 25.7 mm Hg and 85.1 ± 12.4 to 83.7 ± 11.5 mm Hg after 2 minutes of 1.5 mg/kg lidocaine bolus administration, respectively (P = .5 and .7, respectively). Heart rate changed from 80.3 ± 15.1 to 90.6 ± 15.0 beats/min (P = .1), the systolic blood pressure from 127.4 ± 19.5 to 136.3 ± 24.1 mm Hg (P = .4), and the diastolic blood pressure from 76 ± 14.9 to 85.5 ± 15.2 mm Hg (P = .19) at baseline and after administration of 2.5 mg/kg lidocaine bolus, respectively.
Discussion Lidocaine has long been used as an effective agent for the acute treatment of ventricular arrhythmias, particularly after acute myocardial infarction.5,6,8-11,13-15 In the early studies looking primarily at efficacy in ventricular arrhythmias, it was incidentally observed that many patients with atrial fibrillation were converted to sinus rhythm.2,5,7 However, at the time the focus was only on the ventricular rhythm and the observations about atrial fibrillation were never prospectively evaluated. The levels of lidocaine necessary for electrophysiologic effects on the atria are approximately an order of magnitude higher than that needed for ventricular effects.4,5 Consequently, it has been thought that lidocaine has no atrial effect under normal circumstances. Although never studied in human beings, in a dog model lidocaine was found to be effective at converting atrial fibrillation in 101 of 101 episodes of parasympathetically induced atrial fibrillation with reasonable doses.4 Given the less than ideal therapeutic drug options currently available for the treatment of atrial fibrillation, we sought to revisit an old drug for a new use. In this trial, the effect of intravenous lidocaine was evaluated in comparison with placebo, both of which were administered in a double-blind, crossover manner. Neither lidocaine nor placebo was successful in achieving sinus rhythm, whereas transthoracic cardioversion, subsequently performed in 19 of the 20 patients, restored sinus rhythm in 13 patients (68%). There are several possible causes for the failure of lidocaine in our patient population. We gave lidocaine at 2 different concentrations, 1.5 mg/kg and 2.5 mg/kg. Mandel and Bigger4 showed in experimental studies that high concentrations of lidocaine would prolong the effective refractory period in isolated canine and rabbit atrial tissues. We adjusted our dose to that used in this study, anticipating transient levels in the range shown to prolong the effective refractory period. The finding that short duration parasympathetically induced atrial fibrillation in the dog responded to bolus lidocaine suggests that these doses would prolong the atrial effective refractory period sufficiently to terminate atrial fibrillation. Failure in the clinical human population may relate to differences in the nature and duration of atrial fibrillation between the dog and
human being or differences in response of atrial tissue to lidocaine. In the dog study, atrial fibrillation was of short duration and induced parasympathetically. Experimental data suggest that longer atrial fibrillation duration enhances the substrate for the maintenance of atrial fibrillation by shortening the action potential duration and the effective refractory period in the atrial tissue, electric remodeling.10 Action potential duration prolongation induced by lidocaine may be insufficient in the chronic state to terminate atrial fibrillation. Furthermore, spontaneous atrial fibrillation may represent an atrial substrate less susceptible to lidocaine than vagally induced atrial fibrillation. The possibility that bolus lidocaine may be effective in the unusual case of vagal-mediated atrial fibrillation in the human being is not excluded by this study. Because lidocaine levels were not measured after the bolus, it is not known whether adequate levels were actually reached. Although administering lidocaine at higher concentrations may show a better efficacy, given the side effect profile at the high dose already used in this study still higher doses are not likely to be accepted by patients. An incidental finding is that rapid boluses of 1.5 to 2.5 mg/kg of lidocaine, though associated with frequency transient neurologic symptoms, are safe. There were no severe hemodynamic consequences from use of doses as high as 2.5 mg/kg given as a 10-second bolus. In the patient with acute clinical indications for lidocaine, this should provide comfort in rapid treatment.
Limitations The study was conducted in patients with clinical atrial fibrillation without prior screens to choose those most likely to respond to lidocaine. Consequently, lidocaine might be efficacious in subsets of the general atrial fibrillation population, particularly those with acute onset atrial fibrillation. In addition, the fact that this study focused on a relatively small patient population does not rule out a small therapeutic efficacy for lidocaine of <15%.
Summary Lidocaine, though demonstrated in high doses to prolong the atrial refractory period in atrial muscle tissue preparations, and to terminate parasympathetically
induced atrial fibrillation in dogs, is not effective in routine cardioversion of atrial fibrillation in human beings. This result raises questions regarding the appropriateness of using the vagal-mediated atrial fibrillation dog model in therapeutic trials of agents to be used in human atrial fibrillation.
References 1. Kannel W, Abbott R, Savage DD, et al. Epidemiologic features of atrial fibrillation: The Framingham Study. N Engl J Med 1982;306:1018. 2. Baily D, Lehmann M, et al. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation [abstract]. J Am Coll Cardiol 1992;19:41A. 3. Sandercock P, Bramford J, Dennis M, et al. Atrial fibrillation and stroke: prevalence in different types of stroke and influence on early and long-term prognosis. BMJ 1992;305:1460-5. 4. Fagbemi SO, Chi L, Lucchesi BR. Antifibrillatory and profibrillatory actions of selected class I antiarrhythmic agents. J Cardiovasc Pharmacol 1993;21:709-19. 5. Flensted-Jensen E, Sandoe E. Lidocaine as an antiarrhythmic agent. Acta Med Scand 1969;185:297-302. 6. Rodman J, Jelliffe R, Kolb E, et al. Clinical studies with computer assisted initial lidocaine therapy. Arch Intern Med 1984;144:703-9. 7. Rowland M, Thompson P, Guichard A, et al. Disposition kinetics of lidocaine in normal subjects. Ann NY Acad Sci 1971;179: 383-98. 8. Mandel W, Bigger J. Electrophysiologic effects of lidocaine on isolated canine and rabbit atrial tissue. J Pharmacol Exper Ther 1971; 178:81-93. 9. Johnson RG, Goldberger AL, Thurer RL, et al. Lidocaine prophylaxis in coronary revascularization patients: a randomized, prospective trial. Ann Thorac Surg 1993;55:1180-4. 10. Chopra M, Portal R, Aber C. Lignocaine therapy after acute myocardial infarction. BMJ 1969; i:213-6. 11. Coplen S, Antman E, Berlin JA, et al. Efficacy and safety of quinidine therapy for maintenance of sinus rhythm after cardioversion: a metaanalysis of randomized control trials. Circulation 1990;82:1106-16. 12. Malach M, Kostis J, Fischetti. Lidocaine for ventricular arrhythmias in acute myocardial infarction. Am J Med Sci 1969;257:52-60. 13. Grossman J, Lubow L, Frieden J, et al. Lidocaine in cardiac arrhythmias. Arch Intern Med 1968;121:396-401. 14. David D, Lang RM, Neumann A, et al. Parasympathetically modulated antiarrhythmic action of lidocaine in atrial fibrillation. Am Heart J 1990;119:1061-8. 15. Wijffels MCEF, Kirchhof CJHJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. Circulation 1995;92:1954-68. 16. Jewitt D, Kishon Y, Thomas N. Lignocaine in the management of arrhythmias after acute myocardial infarction.” Lancet 1968;i:266-70.
Background Most drugs used for chemical cardioversion of atrial fibrillation have significant proarrhythmia risk and require close monitoring after administration. Lidocaine has few of the proarrhythmic concerns of most antiarrhythmic drugs and, at high bolus doses, prolongs the atrial refractory period well enough to be effective in converting atrial fibrillation to sinus rhythm. This finding has been previously demonstrated in a dog model. We sought to confirm the animal findings in human beings with lidocaine doses of 1.5 to 2.5 mg/kg.
Methods Twenty patients with atrial fibrillation scheduled for elective cardioversion were enrolled in this study. In a randomized, double-blind, crossover study design, each patient received intravenous bolus lidocaine or saline. Patients were observed for 10 minutes after the initial bolus to assess efficacy. The second test drug was then delivered if the first was unsuccessful at cardioversion.
Results All 20 patients received both lidocaine and saline placebo therapy in a crossover manner. None of the 20 patients converted to sinus rhythm with either therapy. The 95% confidence interval for effectiveness of lidocaine in this population was 0% to 14%.
Conclusion In this population of patients referred for elective cardioversion of atrial fibrillation, high-dose bolus lidocaine was ineffective in converting patients to sinus rhythm. Although this study was not sufficiently powered to rule out a low efficacy of lidocaine (<15%) or a higher efficacy in certain subgroups of atrial fibrillation, routine use of lidocaine for this indication is not warranted. (Am Heart J 2000;139:e6.)
Atrial fibrillation is the most common arrhythmia in the United States, affecting 0.4% of the adult population. It accounts for 3.5% of all cardiology hospital admissions and 35% of admissions for arrhythmia. At least 18% of first strokes can be attributed to atrial fibrillation.1,2 To decrease stroke risk, improve hemodynamics, and restore normal sinus node activity, cardioversion of atrial fibrillation is part of common clinical practice.3 Cardioversion is usually accomplished by use of transthoracic shocks with appropriate anesthesia or by pharmacologic means with appropriate follow-up monitoring for proarrhythmia. The availability of an inexpensive agent for conversion of atrial fibrillation without sedation or prolonged observation would allow cardioversion to be attempted in an outpatient clinic setting rather than as a hospital procedure. From the Department of Medicine, University of Washington. Supported by a fellowship award from the American Heart Association, Washington affiliate. Submitted August 9, 1999; accepted January 24, 2000. Reprint requests: Gust H. Bardy, MD, Clinical Coordinating Center, 7900 East Greenlake Dr North, Suite 300, Seattle, WA 98103-4819. E-mail: [email protected] Copyright © 2000 by Mosby, Inc. 0002-8703/2000/$12.00 + 0 4/90/106169 doi:10.1067/mhj.2000.106169
Intravenous lidocaine is a readily available agent familiar to all physicians. It is generally used for ventricular arrhythmias because of its minimal effect on atrial myocardium at low to moderate doses. At higher doses, achievable with rapid bolus administration, lidocaine has the property of prolonging the atrial refractory period, similar to class III agents.12,13,15,16 Such high serum levels cannot be maintained with continuous infusion because of neurologic side effects, but cardioversion should require only transient prolongation of the atrial refractory period, similar to the effect of electrical cardioversion. In fact, in a dog model of atrial fibrillation, bolus lidocaine was effective in terminating all of 101 episodes of atrial fibrillation.4 Consequently, we sought to evaluate the efficacy of high dose intravenous bolus lidocaine for cardioversion of atrial fibrillation in human beings in a prospective, placebo-controlled manner.
Methods Patient population Patients referred for elective electrical or chemical cardioversion of atrial fibrillation were screened for the study. Institutional review board approval was obtained for the study,
Table I. Patient demographics General No. of patients Male Age (y) Height (cm) Weight (kg) Cardiovascular history Coronary artery disease Nonischemic dilated cardiomyopathy Valvular disease Without structural heart disease (lone atrial fibrillation) Noncardiovascular history Hypertension Diabetes mellitus Hypothyroidism Hyperthyroidism Smoker Alcohol use (>1 drink/d) Left atrium size (mm) Atrial fibrillation duration (d) Left ventricular ejection fraction (%) NYHA classification I II III
20 15 62 ± 12 174 ± 4 90 ± 21 8 (40%) 3 (15%) 2 (10%) 7 (35%) 10 (50%) 2 (10%) 2 (10%) 1 (5%) 3 (15%) 6 (30%) 51 ± 6 187 ± 317 47 ± 17 10 (50%) 8 (40%) 2 (10%)
Values are n or n (%).
and all patients had to provide informed written consent for participation. Patients with atrial flutter at the time of cardioversion were excluded. Patients with a history of seizures, liver disease, or sensitivity to lidocaine were also excluded. Of 40 patients approached for the study, 2 were excluded because they spontaneously converted to normal sinus rhythm before randomization, 3 were excluded because they had atrial flutter immediately before the procedure, 3 were excluded because of a history of seizure disorder, and 1 was excluded because of a history of liver disease. Of the remaining 31 patients, 11 declined to participate in the study; thus 20 patients were enrolled.
Lidocaine administration protocol Half of the 20 patients were assigned a lidocaine dose of 1.5 mg/kg and half a dose of 2.5 mg/kg. At the time of cardioversion, syringes with equal volume of lidocaine and saline placebo were prepared and coded in a manner not revealing their contents by the hospital research pharmacy. In a random and blinded manner, an initial treatment with either saline placebo or lidocaine was injected over a 10second period with continuous electrocardiographic and noninvasive blood pressure monitoring for 10 minutes. If patients did not convert to normal sinus rhythm within 10 minutes, they were crossed over to the other arm and the second 10 mL bolus of either saline or lidocaine was infused over a 10-second period. Patients were monitored for an additional 10-minute period. If, after crossover, the patient remained in atrial fibrillation, a transthoracic electrical cardioversion was performed. Continuous single-lead electrocardiographic recordings
were made from 1 minute before administration of each agent to 5 minutes after administration of the last test drug. The primary end point was conversion of atrial fibrillation to normal sinus rhythm within 5 minutes of drug administration. Vital signs were recorded at baseline and at 2 minutes after drug administration to assess the hemodynamic consequence of bolus lidocaine.
Statistical methods The study design was based on the premise that an efficacy of 20% would be required for the acceptance of lidocaine as a useful agent for conversion of atrial fibrillation, assuming the side effects from the bolus were tolerable. An initial sample size of 50 was chosen to provide 90% power to identify a 20% efficacy above placebo at significance of .05. However, after the initial 20 patients were enrolled with no successful conversions in either arm, an interim analysis revealed a 99% certainty that the efficacy would be <20% and a 95% certainty that the efficacy would be <15%. The study was therefore terminated for lack of efficacy.
Results Clinical characteristics Of the 20 patients enrolled, 15 (75%) were men. The mean age was 62 ± 12 years (range 46 to 84 years), the mean height was 174 ± 9 cm (range 157 to 193 cm), and the mean weight was 90 ± 21 kg (range 52 to 150 kg). Ten patients (50%) had a history of hypertension, whereas only 2 patients (10%) had diabetes. Two patients (10%) had hypothyroidism and were receiving replacement therapy, and 1 patient (5%) had a history of treated hyperthyroidism. Only 3 patients (15%) were smokers. One patient (5%) admitted to heavy alcohol use (>3 drinks per day), whereas 5 patients (25%) were moderate drinkers (1 to 2 drinks per day), and 14 patients (70%) drank rarely or not at all. The duration of atrial fibrillation was 187 ± 317 days with a range of 0 to 1140 days. Two patients had postoperative atrial fibrillation. Sixteen patients (75%) were anticoagulated at the time of cardioversion with mean international normalized ratio of 2.77 ± 0.55. All the patients who were anticoagulated had therapeutic international normalized ratios of ≥2.0 for 3 weeks before cardioversion. The 4 patients who were not anticoagulated had atrial fibrillation of <48-hour duration. For 14 patients (70%), this was the first cardioversion; 6 patients (30%) had 1 or more previous successful electrical cardioversions (mean 1.8). Underlying structural heart disease included coronary artery disease in 8 patients (40%), nonischemic dilated cardiomyopathy in 3 (15%), and isolated valvular disease in 2 (10%). The remaining 7 patients (35%) had lone atrial fibrillation without other identifiable underlying cardiac disease. Left ventricular function was normal in 11 patients (55%) and only mildly depressed, with ejection fraction of 40% to 50% in 3 patients (15%). Six patients (30%) had ejection fraction <40%. Functional
class was normal (New York Heart Association [NYHA] class I) in 10 patients (50%); 8 patients (40%) were NYHA class II, and 2 patients (10%) were NYHA class III. No patient had class IV symptoms. Echocardiographic data were available in 18 of the 20 patients. The mean left ventricular end-diastolic dimension was 56 ± 8 mm, and the mean left ventricular endsystolic dimension was 38 ± 11 mm. The mean fractional shortening was 33% ± 12%. There was left atrial enlargement of >40 mm in all but 1 patient, with mean left atrial size of 51 ± 6 mm. Patients demographics are listed in Table I.
Adverse effects All patients in the protocol reported neurologic symptoms from the lidocaine bolus; 3 patients also reported symptoms from placebo. Fifteen patients (75%) noted dysarthria after the lidocaine bolus, 11 patients (55%) noted paresthesia, 3 (15%) had confusion or anxiety, 2 (10%) had shortness of breath or dry mouth each, and 1 patient (5%) had chest pain. Two patients specifically noted that time seemed compressed, feeling only 20 to 30 seconds had elapsed at 6 minutes. Nine patients (45%) noted visual or ocular changes or hallucinations, with one patient, a pediatric ophthalmologist, specifically describing “vertical oscillopsia.” Central nervous system depression and nausea each occurred in 5 of the 10 patients in the high-dose (2.5 mg/kg) group; neither of these was noted in patients in the lowdose (1.5 mg/kg) group. Six of the 10 patients in the low-dose group noted euphoria; this was not noted in any of the patients in the high-dose group. All the symptoms from the lidocaine bolus resolved within 10 minutes. Three patients reported effects during the placebo infusion, 1 with paresthesia, 1 with shortness of breath, and 1 with nausea. The patient with paresthesia from placebo received placebo first in the crossover design, whereas the other 2 patients received lidocaine first.
Cardioversion results In all 20 patients, both lidocaine and placebo failed to convert atrial fibrillation to normal sinus rhythm, even transiently. Transthoracic cardioversion was subsequently performed in 19 of the 20 patients, with 1 patient refusing to proceed to DC cardioversion because of persistent nausea. Sinus rhythm was restored in 13 patients (68%) with a mean of 1.5 shocks. The other 6 patients (32%) remained in atrial fibrillation.
Hemodynamic effects No significant hemodynamic changes were registered after the administration of 1.5 mg/kg and 2.5 mg/kg lidocaine bolus. Documented heart rate was 86.5 ± 14.8 beats/min at baseline and 91.2 ± 10.9 beats/min 2 minutes after administration of 1.5 mg/kg lidocaine bolus (P = .4). The systolic and diastolic blood pressure changed
nonsignificantly from 130.6 ± 22.3 to 136.4 ± 25.7 mm Hg and 85.1 ± 12.4 to 83.7 ± 11.5 mm Hg after 2 minutes of 1.5 mg/kg lidocaine bolus administration, respectively (P = .5 and .7, respectively). Heart rate changed from 80.3 ± 15.1 to 90.6 ± 15.0 beats/min (P = .1), the systolic blood pressure from 127.4 ± 19.5 to 136.3 ± 24.1 mm Hg (P = .4), and the diastolic blood pressure from 76 ± 14.9 to 85.5 ± 15.2 mm Hg (P = .19) at baseline and after administration of 2.5 mg/kg lidocaine bolus, respectively.
Discussion Lidocaine has long been used as an effective agent for the acute treatment of ventricular arrhythmias, particularly after acute myocardial infarction.5,6,8-11,13-15 In the early studies looking primarily at efficacy in ventricular arrhythmias, it was incidentally observed that many patients with atrial fibrillation were converted to sinus rhythm.2,5,7 However, at the time the focus was only on the ventricular rhythm and the observations about atrial fibrillation were never prospectively evaluated. The levels of lidocaine necessary for electrophysiologic effects on the atria are approximately an order of magnitude higher than that needed for ventricular effects.4,5 Consequently, it has been thought that lidocaine has no atrial effect under normal circumstances. Although never studied in human beings, in a dog model lidocaine was found to be effective at converting atrial fibrillation in 101 of 101 episodes of parasympathetically induced atrial fibrillation with reasonable doses.4 Given the less than ideal therapeutic drug options currently available for the treatment of atrial fibrillation, we sought to revisit an old drug for a new use. In this trial, the effect of intravenous lidocaine was evaluated in comparison with placebo, both of which were administered in a double-blind, crossover manner. Neither lidocaine nor placebo was successful in achieving sinus rhythm, whereas transthoracic cardioversion, subsequently performed in 19 of the 20 patients, restored sinus rhythm in 13 patients (68%). There are several possible causes for the failure of lidocaine in our patient population. We gave lidocaine at 2 different concentrations, 1.5 mg/kg and 2.5 mg/kg. Mandel and Bigger4 showed in experimental studies that high concentrations of lidocaine would prolong the effective refractory period in isolated canine and rabbit atrial tissues. We adjusted our dose to that used in this study, anticipating transient levels in the range shown to prolong the effective refractory period. The finding that short duration parasympathetically induced atrial fibrillation in the dog responded to bolus lidocaine suggests that these doses would prolong the atrial effective refractory period sufficiently to terminate atrial fibrillation. Failure in the clinical human population may relate to differences in the nature and duration of atrial fibrillation between the dog and
human being or differences in response of atrial tissue to lidocaine. In the dog study, atrial fibrillation was of short duration and induced parasympathetically. Experimental data suggest that longer atrial fibrillation duration enhances the substrate for the maintenance of atrial fibrillation by shortening the action potential duration and the effective refractory period in the atrial tissue, electric remodeling.10 Action potential duration prolongation induced by lidocaine may be insufficient in the chronic state to terminate atrial fibrillation. Furthermore, spontaneous atrial fibrillation may represent an atrial substrate less susceptible to lidocaine than vagally induced atrial fibrillation. The possibility that bolus lidocaine may be effective in the unusual case of vagal-mediated atrial fibrillation in the human being is not excluded by this study. Because lidocaine levels were not measured after the bolus, it is not known whether adequate levels were actually reached. Although administering lidocaine at higher concentrations may show a better efficacy, given the side effect profile at the high dose already used in this study still higher doses are not likely to be accepted by patients. An incidental finding is that rapid boluses of 1.5 to 2.5 mg/kg of lidocaine, though associated with frequency transient neurologic symptoms, are safe. There were no severe hemodynamic consequences from use of doses as high as 2.5 mg/kg given as a 10-second bolus. In the patient with acute clinical indications for lidocaine, this should provide comfort in rapid treatment.
Limitations The study was conducted in patients with clinical atrial fibrillation without prior screens to choose those most likely to respond to lidocaine. Consequently, lidocaine might be efficacious in subsets of the general atrial fibrillation population, particularly those with acute onset atrial fibrillation. In addition, the fact that this study focused on a relatively small patient population does not rule out a small therapeutic efficacy for lidocaine of <15%.
Summary Lidocaine, though demonstrated in high doses to prolong the atrial refractory period in atrial muscle tissue preparations, and to terminate parasympathetically
induced atrial fibrillation in dogs, is not effective in routine cardioversion of atrial fibrillation in human beings. This result raises questions regarding the appropriateness of using the vagal-mediated atrial fibrillation dog model in therapeutic trials of agents to be used in human atrial fibrillation.
References 1. Kannel W, Abbott R, Savage DD, et al. Epidemiologic features of atrial fibrillation: The Framingham Study. N Engl J Med 1982;306:1018. 2. Baily D, Lehmann M, et al. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation [abstract]. J Am Coll Cardiol 1992;19:41A. 3. Sandercock P, Bramford J, Dennis M, et al. Atrial fibrillation and stroke: prevalence in different types of stroke and influence on early and long-term prognosis. BMJ 1992;305:1460-5. 4. Fagbemi SO, Chi L, Lucchesi BR. Antifibrillatory and profibrillatory actions of selected class I antiarrhythmic agents. J Cardiovasc Pharmacol 1993;21:709-19. 5. Flensted-Jensen E, Sandoe E. Lidocaine as an antiarrhythmic agent. Acta Med Scand 1969;185:297-302. 6. Rodman J, Jelliffe R, Kolb E, et al. Clinical studies with computer assisted initial lidocaine therapy. Arch Intern Med 1984;144:703-9. 7. Rowland M, Thompson P, Guichard A, et al. Disposition kinetics of lidocaine in normal subjects. Ann NY Acad Sci 1971;179: 383-98. 8. Mandel W, Bigger J. Electrophysiologic effects of lidocaine on isolated canine and rabbit atrial tissue. J Pharmacol Exper Ther 1971; 178:81-93. 9. Johnson RG, Goldberger AL, Thurer RL, et al. Lidocaine prophylaxis in coronary revascularization patients: a randomized, prospective trial. Ann Thorac Surg 1993;55:1180-4. 10. Chopra M, Portal R, Aber C. Lignocaine therapy after acute myocardial infarction. BMJ 1969; i:213-6. 11. Coplen S, Antman E, Berlin JA, et al. Efficacy and safety of quinidine therapy for maintenance of sinus rhythm after cardioversion: a metaanalysis of randomized control trials. Circulation 1990;82:1106-16. 12. Malach M, Kostis J, Fischetti. Lidocaine for ventricular arrhythmias in acute myocardial infarction. Am J Med Sci 1969;257:52-60. 13. Grossman J, Lubow L, Frieden J, et al. Lidocaine in cardiac arrhythmias. Arch Intern Med 1968;121:396-401. 14. David D, Lang RM, Neumann A, et al. Parasympathetically modulated antiarrhythmic action of lidocaine in atrial fibrillation. Am Heart J 1990;119:1061-8. 15. Wijffels MCEF, Kirchhof CJHJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. Circulation 1995;92:1954-68. 16. Jewitt D, Kishon Y, Thomas N. Lignocaine in the management of arrhythmias after acute myocardial infarction.” Lancet 1968;i:266-70.