A Prospective Placebo Controlled Randomized Study of Caffeine in Patients with Supraventricular Tachycardia Undergoing Electrophysiologic Testing ROBERT LEMERY, M.D., AMANDA PECARSKIE, B.Sc., JORDAN BERNICK, M.Sc., KATHRYN WILLIAMS, B.Sc., M.S., and GEORGE A. WELLS, Ph.D. From the Division of Cardiology, University of Ottawa Heart Institute, Ottawa, Ontario, Canada
Study of Caffeine in Patients with Supraventricular Tachycardia. Introduction: Patients with cardiac arrhythmias are generally instructed to avoid caffeine intake. A comprehensive evaluation of the electrophysiological effects of caffeine on atrial and ventricular tissues in humans has not previously been performed. Methods and Results: Eighty patients (31 men, mean age 49 ± 14 years) with symptomatic supraventricular tachycardia (SVT) undergoing an electrophysiologic study (EPS) prior to catheter ablation were randomized to receive oral caffeine or placebo. Caffeine at a dosage of 5 mg/kg (moderate intake) or placebo tablets were administered orally at a mean time of 57 ± 13 minutes prior to the EPS. The median (IQR) caffeine level in patients receiving caffeine was 7.4 µg/mL (4.7–8.7), as compared with 0.15 (0.00–0.61) in patients receiving placebo, P < 0.0001. Caffeine was associated with a significant increase in resting systolic and diastolic blood pressures as compared with placebo, while the resting heart rate was not significantly different between both groups. Caffeine was not associated with significant effects on the effective refractory period of the atrium or ventricle, as well as on AV node conduction. SVT was induced in all but 3 patients; there was no significant difference between groups receiving placebo or caffeine on SVT inducibility or the cycle length of induced tachycardias. Conclusions: Caffeine, at moderate intake, was associated with significant increases in systolic and diastolic blood pressures, but had no evidence of a significant effect on cardiac conduction and refractoriness. Furthermore, no effect of caffeine on SVT induction or more rapid rates of induced tachycardias was found. (J Cardiovasc Electrophysiol, Vol. 26, pp. 1-6, January 2015) arrhythmias, atrial fibrillation, caffeine, electrophysiology testing, supraventricular tachycardia Caffeine has become widely available for consumption, with potentially large quantities associated with energy drinks.1,2 In clinical practice, patients with a history of cardiac arrhythmias are typically instructed by their physician to discontinue caffeine intake.3,4 There have been few reports of invasive electrophysiologic evaluation of the effects of caffeine in animals5-7 and patients.8-10 Studies in cardiac patients or volunteers reported electrophysiological effects on atrial, AV node, or ventricular refractoriness that suggested the potential for caffeine intake to be detrimental in patients with supraventricular tachycardia (SVT) or other cardiac arrhythmias.8-10 The objective of this prospective randomized placebo controlled study was to determine if caffeine in
No disclosures. Presented in part at the Annual Scientific Sessions of the American Heart Association, November 2013, Dallas, Texas. Clinical Trial Registration: ClinicalTrials.gov (Unique identifier: NCT 02095405). Address for correspondence: Robert Lemery, M.D., University of Ottawa Heart Institute, 40 Ruskin, Ottawa, Ontario, Canada K1Y-4W7. Fax: 613761-4407; E-mail: [email protected]
Manuscript received 12 May 2014; Revised manuscript received 22 July 2014; Accepted for publication 25 July 2014. doi: 10.1111/jce.12504
moderate amounts was associated with cardiac electrophysiology effects. Methods Study Population The study group includes 80 patients with symptomatic SVT who were scheduled to undergo radiofrequency ablation. In all patients, a rhythm strip or 12-lead ECG documentation of SVT was obtained prior to performing the electrophysiologic study (EPS) at the time of the ablation procedure. The exclusion criteria consisted of intolerance to caffeinated beverages, as well as use of medications that may react with caffeine through the cytochrome P450 1A2 pathway (CYP1A2). In women of childbearing age, a negative pregnancy test was required prior to the EPS. The clinical and arrhythmia characteristics of both groups of patients are described in Table 1. Clinically, only 15/80 patients (19%) reported a possible relation between caffeine intake and occurrence of palpitations or tachycardia. The most frequent precipitating factors associated with the onset of SVT were none with spontaneous occurrence in 77%, exercise-induced in 25%, and bending over in 25%. During SVT, symptoms typically included palpitations/tachycardia in 99%, chest pain in 30%, nausea in 30%, and presyncope or syncope in 19%. Termination of tachycardia occurred spontaneously during lying down, resting, or deep breathing in 65%, specifically by valsalva maneuvers in 51%, while ER
Journal of Cardiovascular Electrophysiology
Vol. 26, No. 1, January 2015
TABLE 1 Clinical and Characteristics of Tachycardia in Both Groups of Patients Caffeine Group Placebo Group (N = 40) (N = 40) P Value Age (years) Median Interquartile range Female Sex (%) Hypertension – No. (%) Height (cm) Median Interquartile range Weight (kg) Median Interquartile range Body Mass Index Median Interquartile range Symptoms of SVT (years) Median Interquartile range Episodes of SVT per month Median Interquartile range Previous Emergency Room Visits Median Interquartile range Daily Caffeine Consumption (cups) Median Interquartile range
49 39-62 21 (52.5%) 5 (12.5%)
50 42-56 28 (70%) 10 (25%)
For each analysis, there was no significant difference between values for caffeine as compared with placebo.
visits for tachycardia termination with adenosine occurred in 57%. A history of smoking was documented in 13/80 patients (16%), including 5 patients in the caffeine group and 8 patients in the placebo group. Study Design The study was designed to determine if caffeine had any significant electrophysiology effects on cardiac refractoriness and conduction. Immediately prior to entering the electrophysiology laboratory, patients randomly received oral tablets of caffeine or placebo, consisting of 5 mg/kg, rounded to the highest concentration amount. Following insertion of catheters, programmed electrical stimulation was performed following the same sequence of atrial and ventricular stimulation in all patients. Following completion of the protocol, if SVT was not induced, isoproterenol was administered to induce SVT. Following completion of this protocol and the study, patients underwent radiofrequency ablation under standard clinical care and practice.
electrolytes, creatinine, complete blood count, and an ECG were obtained. Patients also consented for two 10 cc blood samples for analysis of caffeine blood levels, to be taken at the time of starting the EPS. Patients answered a questionnaire about their clinical history of SVT, previous emergency room visits, effects of caffeine substances on arrhythmia occurrence, and treatment prior to their scheduled ablation procedure. Electrophysiologic Study Patients underwent electrophysiologic testing in the fasting postabsorptive state. Caffeine or placebo was administered just prior to entering the electrophysiology laboratory. Conscious sedation consisted of administration of midazolam and fentanyl. Catheters were introduced from the right femoral vein and the right internal jugular vein under fluoroscopic guidance to the high RA, RV apex, AV node, and coronary sinus. Baseline variables of blood pressure, heart rate, and intracardiac measurements were obtained while the patient was resting in the supine position prior to performing intracardiac stimulation. Programmed electrical stimulation was performed at twice diastolic threshold according to standard protocol,11,12 at 600 and 500 milliseconds, and when needed at 400 milliseconds. Sinus node recovery times were obtained following pacing from the high right atrium. This was followed by determination of the effective refractory period (ERP) of the RV, and 1:1 conduction retrograde over the AV node, and when present of the fast pathway, slow pathway, and accessory pathway. The same protocol was then followed to determine the ERP and 1:1 conduction when pacing from the high RA and distal coronary sinus. Finally, rapid atrial pacing to 2:1 atrial capture or to 250 milliseconds was performed. For ERP measurements, during stable pacing of 8 beats, premature stimuli from the atrium or ventricle were delivered during diastole, decrementally by 10 milliseconds until refractoriness (ERP, the longest coupling interval for which the premature impulse failed to conduct to the tissue). The minimum coupling interval during delivery of extrastimuli was 180 milliseconds. Endpoints The primary endpoint of the study consisted of obtaining the ERP of the RA and RV. The secondary endpoints consisted of obtaining the ERP of the AV node and CS, antegrade and retrograde AV node conduction block, method of tachycardia inducibility, cycle length of induced tachycardia in the baseline state or following an infusion of isoproterenol, and correlation of administered amounts of caffeine or placebo with caffeine blood levels. The results were analyzed following completion of the entire study.
The study was performed at the University of Ottawa Heart Institute. The Human Ethics Review Board approved the study protocol, and written informed consent was obtained from all patients. Throughout the conduct of the trial, there were never any changes made to the submitted and approved protocol. A computer-generated randomization table was used for randomization. Patients were seen in clinic, or the emergency room, or the preadmission unit prior to randomization. An electrocardiogram, blood tests including
In order to detect a 10% reduction in the primary endpoint (refractory period) under the experimental group “caffeine” at alpha = 0.05 (2-sided) and 80% power, a sample size of 80 patients (40 per group, caffeine vs. no caffeine) was needed. A dropout rate of 5% was included in the sample size calculation. A 10% reduction in the refractory period was identified as clinically important. In the nonrandomized study of caffeine in volunteers and patients, significant mean decreases of 10-20 milliseconds on atrial or ventricular
Lemery et al. Study of Caffeine in Patients with Supraventricular Tachycardia
refractoriness were documented.9 Josephson has reported that measurements of conduction and refractoriness are generally reproducible (± 10%) over a period of several hours for atrial and ventricular muscle,12 while maintaining comparable stimulus strength. Continuous outcomes are presented as mean ± SD, or median and interquartile range, and categorical outcomes as proportions. Study groups were compared using the MannWhitney U test for continuous outcomes and Fisher’s exact test for categorical outcomes. A regression analysis was conducted to assess the relationship between caffeine and ERP. A 2-sided P value of less than 0.05 was considered to indicate statistical significance. The statistical tests were performed with SPSS software, version 9.3. Results Tablets of caffeine were administered at a mean time of 54 ± 11 minutes prior to starting the EPS, as compared with 59 ± 15 minutes for patients receiving placebo (P = 0.06). The time interval to complete the study protocol, including administration of isoproterenol when necessary to induce SVT, was 60 ± 32 minutes in patients receiving caffeine, as compared with 65 ± 27 minutes in the placebo group (P = 0.48). None of the patients had induction of ventricular tachycardia or ventricular fibrillation during programmed electrical stimulation. Resting Blood Pressure, Heart Rate, and Electrophysiologic Intervals In the supine resting state, the systolic and diastolic blood pressures were significantly higher in patients receiving caffeine, while the resting heart rate was not significantly different between both groups (Table 2). The baseline resting electrophysiologic intervals were not significantly different between groups. Measurements of Intracardiac Refractoriness and Conduction There were no significant differences in values obtained when measuring sinus node recovery times, or corrected sinus node recovery times between both groups (Table 3). The right ventricular ERP and VA conduction time were not significantly different between groups. The right atrial ERP, the AV node ERP, and the AV node conduction time were not significantly different between patients receiving caffeine as compared with patients receiving placebo. When pacing the CS, the left atrial ERP was also similar between groups. However, when measuring the ERP of the atria in the caffeine group, there was a slight shift of the range of values to a lower value as compared with placebo, but the median and mean values were not significantly different. Rapid Atrial Pacing Rapid atrial pacing to 2:1 capture or to 250 milliseconds was performed in 63 patients (79%); in the remaining patients, rapid atrial pacing was not performed either due to repeated SVT induction, or because the physician had completed the baseline protocol and proceeded with ablation. Induction of atrial fibrillation or atrial flutter occurred in 9 patients (22%) in the caffeine group (rapid pacing was performed in 30/40 patients), as compared with 14 patients
TABLE 2 Measurements of Caffeine Blood Levels, Blood Pressure, Heart Rate, and Resting Electrophysiologic Intervals Caffeine Group Placebo (N = 40) (N = 40) Caffeine Blood Level (µg/mL) Median Interquartile range Blood Pressure (mmHg) Systolic Median Interquartile Range Diastolic Median Interquartile range Heart Rate (bpm) Median Interquartile range Cycle length in sinus rhythm (milliseconds) Median Interquartile range AH (milliseconds) Median Interquartile range HV (milliseconds) Median Interquartile range QRS (milliseconds) Median Interquartile range QT (milliseconds) Median Interquartile range