Combining Monophasic Action Potential Recordings With Pacing to Demonstrate Delayed Afterdepolarizations and Triggered Arrhythmias in the Intact Heart
Value of Diastolic Slope
Background In the intact heart, methodological difficulties hamper the direct visualization of delayed afterdepolarizations (DADs) responsible for triggered arrhythmias. Therefore, we tested the hypothesis that a combination of pacing and the recording of a monophasic action potential (MAP) could facilitate the recognition of ouabain-induced DADs and triggered arrhythmias by demonstrating an increase in the diastolic baseline slope (dV/dT) of the MAP recording at the end of a pacing train.
Methods and Results In anesthetized dogs with chronic atrioventricular block, a right ventricular endocardial MAP was recorded during (1) control (n=11), (2) 15 to 45 minutes after administration of ouabain (45±10 μg/kg, n=11), (3) 10 minutes after administration of lidocaine (3 mg/kg, n=5), and (4) during lidocaine washout (n=3). Pacing was performed with the MAP catheter. Also, the protocol was performed in 3 dogs with conducted sinus rhythm during control and ouabain circumstances. During control, the slope value was 2±2 mV/s (mean±SD), the incidence of DADs after the stimulation train was 6%, and no ventricular tachycardias (VTs) were induced in dogs with atrioventricular block. During ouabain administration, the slope and DAD incidences increased to, respectively, 26±14 mV/s and 74% (P<.05 for both). VTs were induced frequently. Lidocaine prevented VT induction by decreasing the slope and the incidence of DADs. This effect disappeared after lidocaine washout. During conducted sinus rhythm, similar results were found.
Conclusions By combining pacing and MAP recordings, the diastolic slope observed on MAP recordings in ouabain-intoxicated hearts can be used as a marker for DADs and triggered arrhythmias. This finding may be helpful in identifying triggered activity in the intact heart.
The arrhythmogenic mechanism triggered activity consists of two subgroups: EADs or DADs.1 2 3 4 5 Although many attempts have been made to identify DADs as the cause of clinical tachycardias,6 7 8 9 10 11 12 13 14 15 16 their relevance is still poorly understood. A direct approach could be the visualization of DADs through the recording of endocardially placed MAPs. This approach has been successfully applied in animals during (nonsustained) VTs17 18 19 20 and, more recently, during atrial tachycardias in patients.16 However, methodological difficulties still hamper the general use of this technique.21 22 To elucidate DADs as the substrate of tachycardia initiation, we hypothesized, on the basis of the frequency dependence of DADs, that their occurrence is indicated by the presence of an increase in the baseline slope (dV/dt) in the MAP recordings of the last beats of pacing. This increase in slope has been observed in action potentials recorded in isolated tissue5 before induction of (triggering) DADs. Although the slope of the ascending limb of DADs has been calculated,23 24 the slope during pacing has never been quantified. If our hypothesis is correct, it might provide a tool to identify triggered activity in the intact heart, without the necessity to induce VT.
All experiments were performed in accordance with the “Guiding Principles in the Care and Use of Animals” as approved by the American Physiological Society and under the regulations of the Committee for Experiments on Animals of the University of Limburg, Maastricht, The Netherlands.
The studies were performed in 11 anesthetized dogs (mean body weight, 25±5 kg) of either sex. Complete atrioventricular block was made to exclude the interference of conducted sinus beats on ventricular arrhythmias.25 During the atrioventricular block operation, an epicardial screw-in electrode (Bakken Research Center) was placed on the apex of the left ventricle. The experiments were performed at least 2 weeks after this operation. Also, 3 anesthetized dogs with normal conducted sinus rhythm were studied.
Stimulation was done with a programmable stimulator that has a synchronizing circuit. Six surface ECG leads and one MAP recording were simultaneously registered on a monitor and stored on an optical disk. Drugs were administered through a cannula in a cephalic vein.
Experiments were performed with the animals under anesthesia and with sterile conditions. Premedication (0.2 mL/kg) was given intramuscularly with a drug mixture containing 10 mg oxycodan HCl, 1 mg acepromazine, and 0.5 mg atropine sulfate per milliliter. An antibiotic (1000 mg ampicillin [Amfipen] Gist Brocades) was given preoperatively (intravenously) and postoperatively (subcutaneously). Anesthesia was induced by sodium pentobarbital (20 mg/kg IV). The dogs were artificially ventilated (Pulmomat respirator, Dräger) through a cuffed endotracheal tube with a mixture of oxygen, nitrous oxide, and halothane (vapor concentration, 0.5% to 1.0%). Ventilation was adjusted on the basis of continuous monitoring of the carbon dioxide concentration of the expired air. A thermal mattress was used to maintain body temperature around 37°C during the experiments.
Recording of MAPs
To register an endocardial MAP signal, a quadripolar contact electrode26 27 (Steerable Franz Combination Catheter, EP Technologies) was placed endocardially in the right ventricle. The 7F MAP catheter was introduced under fluoroscopy through the femoral or external jugular vein. The MAP signals were amplified with a customized isolated DC-coupled differential amplifier with a 20-mV calibration pulse. The offset of this amplifier is variable and can be adjusted to the recorded signal. The MAP signals are sampled at a rate of 1 kHz per signal.
MAP phases were divided according to the definitions used for the transmembrane action potential. Amplitude was defined as the difference between phases 4 and 2 of the signal. The minimally accepted amplitude of the MAP was 15 mV.17 18 If the MAP had a lower amplitude under baseline conditions, the catheter was moved to an adjacent side. To ensure that the recording place of the catheter could be repeatedly reached throughout the experiment, the catheter was placed at the same location at least three times before the experiment was started. This was achieved using guidance from the fluoroscopy monitor and through pace mapping with the MAP catheter.
To avoid artifacts during phase 4 of the action potential,27 rigorous attention was paid to record stable diastolic potentials. Furthermore, the MAP had to have a constant configuration and a smooth shape during control circumstances.
A protocol was used that we routinely use in conscious dogs.19 28 Pacing was performed from the MAP catheter with a strength of twice diastolic threshold and a pulse width of 2 milliseconds and consisted of three stimulation trains with an interstimulus interval of 300 milliseconds and a duration of 4, 10, or 20 seconds. This resulted in three pacing trains in every dog, both during control (11 atrioventricular block dogs and 3 sinus rhythm dogs) and during each intervention. Between these pacing trains, sufficient time was allowed for the heart to regain its prepacing rhythm. After control pacing, a bolus of ouabain (in relation to body weight28 ; mean, 45±10 μg/kg) was administered intravenously over 1 minute. Ten minutes later, this was followed by continuous infusion of ouabain29 (mean, 0.08±0.02 μg/kg per minute) throughout the experiment. Pacing was resumed 15 minutes after the bolus of ouabain.28
Five of the 11 atrioventricular block dogs received an intravenous bolus dose of lidocaine (3 mg/kg in 2 minutes) to reverse the ouabain effects and to exclude possible artifacts in the MAP recordings. The pacing protocol was repeated within 10 minutes after lidocaine administration. Finally, in 3 dogs, the pacing protocol was repeated once more after a washout period of lidocaine.
In 5 atrioventricular block dogs, the described pacing protocol was expanded with three pacing trains that had an interstimulus interval of 400 milliseconds and the same durations (4, 10, and 20 seconds).
For off-line analysis of the recorded signals, a customized ecgview program was used, which contains a special module for the analysis of MAP signals on a beat-to-beat basis. It can calculate the amplitude, duration, and slope (dV/dt) of a selected MAP signal. Slope was calculated as the mean rate of rise (dV/dt) of the ascending limb of the DAD22 23 or as the mean change in diastolic potential between the last five paced beats (Fig 1⇓). The slope calculation is based on the least-squares regression line calculation of the measured points in a selected time window. The begin and end points of this time window are manually adjusted.
To relate the diastolic baseline slope to ouabain-induced DADs and triggered arrhythmias, the following parameters were quantified: (1) absolute value of the slope during pacing, (2) occurrence of DADs directly after pacing, (3) amplitude of the DAD, (4) induction, number, and coupling interval of ectopic beats, and (5) induction of VT. The presence of a DAD after pacing was defined as the occurrence of a depolarizing afterpotential with an ascending and a descending limb that begins after the complete repolarization (phase 3) of the last paced MAP.2 In some cases, the DAD follows an initial phase of hyperpolarization.2 This hyperpolarization was excluded in the analyses of DAD and in the calculation of slope, which were both determined with phase 4 of the preceding action potential as baseline. Analyses of both the amplitude and coupling interval of DADs were performed after pacing conform to methods that have been previously described.23 VT was defined as more than five ectopic beats with a coupling interval shorter than the last spontaneous prepacing idioventricular interval.28
To ensure accuracy of our analysis and guarantee uniform interpretation of the results, all experiments were analyzed twice. One of these analyses was done by an independent observer without knowledge of the experimental circumstances.
All data are presented as mean±SD. ANOVA followed by Bonferroni’s t test (for more than two-group analysis), two-tailed Student’s t test for unpaired events (for two-group comparison), and χ2 analysis was used to determine statistical significance.
Experiments During Chronic Atrioventricular Block
Control Versus Ouabain
An almost flat slope was present during control pacing (Fig 2⇓); the mean calculated value was 2±2 mV/s, and DADs were induced in 6% of the pacing trains (Table 1⇓). These DADs had a mean amplitude of 0.4±0.1 mV. The postpacing incidence of ectopic beats was 18%, with a mean±SD number of 2±1 and a mean coupling interval of 517±186 milliseconds. VTs were not induced.
Ouabain administration resulted in a nonsignificant decrease in the spontaneous idioventricular cycle length (from 1432±377 to 1343±229 milliseconds) and a significant decrease in MAP amplitude (32±13 to 26±13 mV; P<.05). Slope increased to 26±14 mV/s (P<.05), and there was an increase in incidence and amplitude of DADs (74% and 1.0±0.6 mV; P<.05 for both; Fig 3⇓). Furthermore, pacing increased the incidence of ectopic beats (85%; P<.05) and VT (33%; P<.05; Table 1⇑). An example of the induction of VT is given in Fig 4⇓ . The number of ectopic beats increased to 23±47, whereas their coupling interval decreased to 481±282 milliseconds. Neither was significantly different from control.
Lidocaine and Lidocaine Washout Measurements
Lidocaine decreased the MAP amplitude to 18±7 mV (P<.05). However, during the washout period of lidocaine, the amplitude rose to 24±3 mV (P<.05). Lidocaine increased the prepacing idioventricular cycle length to 1861±439 milliseconds (P<.05), which reversed during washout.
Lidocaine decreased the diastolic slope to 4±2 mV/s (P<.05), reduced the incidence of DADs to 57% (P<.05), reduced their mean amplitude to 0.7±0.4 mV (P<.05), and completely prevented the occurrence of VT (P<.05; Fig 4⇑  and Table 1⇑). The incidence of ectopic beats decreased to 64% (P<.05 compared with ouabain); the number of induced beats decreased to 2±2, with an increase in the coupling interval to 802±292 milliseconds (P<.05). After washout of lidocaine, all parameters increased significantly: slope to 24±7 mV/s, induced VTs to 55%, DADs to 77% with a mean DAD amplitude of 1.1±0.2 mV, and incidence of ectopic beats to 100% (Table 1⇑).
During control, the measured parameters did not differ when the two interstimulus intervals were compared in the 5 dogs in which both protocols were performed. The control slope was 2±1 mV/s, and no DADs or VTs were induced in these dogs.
During ouabain, the slope and the incidence of DADs and VTs increased more markedly after pacing with an interstimulus interval of 300 milliseconds (Table 2⇓). Also, the incidence of the ectopic beats after a 400-millisecond pacing train was lower, and their coupling interval was longer than after a 300-millisecond pacing train, but this did not reach statistical significance.
However, the amplitude of induced DADs was higher after the 400-millisecond stimulation train.
Experiments During Sinus Rhythm
Also, during sinus rhythm, the slope remained flat (3±1 mV/s) during control circumstances (Fig 5⇓ ). There were no DADs induced. After ouabain, both of these parameters increased significantly: slope to 17±4 mV/s and DAD incidence to 84% (Fig 5⇓ ). The DAD coupling interval and amplitude were 241±13 milliseconds and 1.0±0.5 mV, respectively. The sinus cycle length was 497±12 milliseconds, which decreased slightly after ouabain administration, to 460±50 milliseconds. MAP amplitude showed a nonsignificant decrease after ouabain (34±20 versus 27±10 mV).
The results of the present study indicate that the diastolic slope at the end of a pacing train can (1) be a marker for the ability to induce DAD-related triggered activity TA, (2) possibly be used as a diagnostic tool to identify DAD-dependent TA, and (3) be of value to quantitatively assess the effect of (pharmacological) interventions on DAD-dependent TA.
Mechanism of Arrhythmias During Ouabain Intoxication
Partial inhibition of the Na+/K+ pump by ouabain causes an increase in [Na+]i concentration that results in an increased [Ca2+]i.3 30 31 This [Ca2+]i overload causes the sarcoplasmic reticulum to release Ca2+ into the cytosol, eliciting a transient inward current (Iti)3 30 31 32 33 that is responsible for DADs. Proposed mechanisms underlying the Iti are a Ca2+-activated nonspecific channel and the Na+/Ca2+ exchanger.32 33
Identification of DAD-Dependent Triggered Activity
The identification of DAD-related TA in patients is difficult because most diagnostic methods are indirect or not available for clinical use. Indirect diagnostic features that have been used include the behavior of DAD-dependent atrial and VTs in relation to pacing9 10 11 16 and termination of DAD-dependent arrhythmias with specific pharmacological probes.13 14 15 16 34 35 36 A more direct method that has been used in the intact heart could be the registration of afterdepolarizations by MAP signals.16 17 18 19 20 However, these recordings should be interpreted with caution21 22 because MAP recordings are not completely identical to transmembrane measurements. Moreover, the MAP technique still has some practical problems, such as MAP stability, artifact registration, and selection of the recording site with regard to site of DAD origin. To minimize the influence of these problems in registering DADs, the following criteria should be fulfilled: the MAP catheter should be placed by an expert to record a stable MAP with an appropriate shape, the equipment used to record and analyze MAP signals must be of high quality, and DADs should be suppressed by an intervention.
The model used in the present study has been extensively tested and used in our laboratory in conscious dogs.19 28 34 37 To introduce the MAP catheter, general anesthesia was used. Otherwise, the model remained the same. Pacing during progressive ouabain intoxication eventually leads to (1) the occurrence of ectopic beats, with (2) shortening of their coupling interval and (3) the induction of VT. In these experiments, however, we did not wait until VT was induced; this allowed us to study the relation between the diastolic slope and the occurrence of DADs, the induction of ectopic beats, and VT.
The Behavior of the Slope Under Different Circumstances
The recognition of DAD-dependent TA without the necessity to induce sustained arrhythmias can be an important advantage when using the MAP technique in clinical practice. It is known that pacing can provoke DADs under specific conditions because DADs are rate dependent.4 5 17 18 23 24 28 34 37 As described in vitro during ouabain administration, the slope in between stimulated beats becomes progressively more steep and the last paced beat is frequently followed by a subthreshold DAD or a triggered action potential.4 5 17 18 19 20 23 24 We studied this relation in dogs with atrioventricular block. In this way, interference of the conducted sinus beats on ventricular arrhythmias is excluded.25 However, to ensure the applicability of the diastolic slope during the much faster sinus rate, the experiments were also performed in dogs with normal atrioventricular conduction.
Our data demonstrate that in the intact globally intoxicated heart, the slope between paced stimuli might represent the presence of DADs. During control circumstances, the slope is relatively flat, although a small increase was observed during pacing. This sometimes resulted in the induction of DADs and ectopic beats but never in the generation of a series of triggered beats or VT. In conscious dogs with atrioventricular block and biventricular hypertrophy, the induction of these ectopic beats by pacing has been described previously.37
During ouabain administration, pacing resulted in a markedly increased slope that was accompanied by a higher incidence of DADs, ectopic beats, and VT. The difference in the occurrence of ventricular ectopy and occurrence of DADs (85% versus 74%) might be explained by a discrepancy between the MAP recording site and site of origin of the DAD. This phenomenon will be studied more intensively in future research.
Lidocaine is known to suppress DADs and related arrhythmias in vitro23 38 and in the intact heart.36 Although it is not a DAD-specific drug, we have used lidocaine because of its practical advances (eg, relatively short half-life and no complications). Lidocaine completely suppressed VT, which was accompanied by a decrease in slope, incidence of DADs, and ectopic beats. These changes disappeared after washout, supporting our hypothesis that the slope is a useful indicator of DADs and TA. The decreased MAP amplitude during lidocaine might be related to negative inotropic effects, considering the increase in MAP amplitude after lidocaine washout.39
In another series of experiments in which the behavior of the slope during VT was studied, we used ryanodine as the specific pharmacological probe. Thirteen minutes after administration of ryanodine (10 μg/kg in 10 minutes), all tachycardias terminated (n=5); this was accompanied by a significant decrease in slope (personal observations).
Influence of Paced Cycle Length
If the diastolic slope is representative for DADs and TA, the slope should be influenced by the interstimulus interval.4 5 10 12 17 18 23 24 Under control circumstances, no difference was seen in the slope at different pacing rates. After ouabain administration, the slope behaved in accordance with our hypothesis that it would be less pronounced at slower pacing rates. A relation was found among pacing rate and incidence of DADs, the occurrence of ectopic beats, and the induction of VT. The only discrepancy is the amplitude of the first DAD postpacing, which is higher after the 400-millisecond stimulation train. This is, however, a difference caused by analysis because a direct comparison of the two dogs that showed DADs after stimulation with both frequencies revealed a higher DAD amplitude after the 300-millisecond stimulation trains.
Another possible explanation for a difference in DAD amplitude could have been the occurrence of two DADs after 300-millisecond pacing and one DAD after pacing with a 400-millisecond interstimulus interval. Although the occurrence of two DADs after the 300-millisecond pacing trains was seen twice in our total group of experiments, this phenomenon was not observed in this series of five dogs.
In the present study, slope was measured in a randomly placed right ventricular MAP recording without consideration of the relevance of the place of origin of the DADs in relation to the registration site. Pacing was always performed from the MAP catheter, which also was used to register the DADs. A stable MAP signal with a certain amplitude was considered to be of greater importance than a search for specific sites in which DADs occur.40 We used a globally intoxicated dog heart and do not know to what extent our registrations of DADs are a true manifestation of DADs at the site of origin. Other experiments are necessary to validate our approach during the pathological circumstances of [Ca2+]i overload, possibly occurring in more localized areas. It should also be evaluated in patients with idiopathic VT and atrial tachycardia.16
In summary, the diastolic slope of the MAP recording during pacing might represent a parameter with which to evaluate the presence of DADs and TA in the intact heart. This could be demonstrated by measuring the diastolic slope during a simple pacing train in circumstances that are suspect for the occurrence of DADs, such as during ouabain intoxication. This suggests that the slope could possibly be used as a quantitative marker with which to evaluate interventions on DAD-dependent arrhythmias in both animal heart and humans.
Selected Abbreviations and Acronyms
|MAP||=||monophasic action potential|
This study was supported by a grant from the Wynand M. Pon Foundation, The Netherlands. We thank the Bakken Research Center, Maastricht, The Netherlands, for supplying the electrodes and the University of Limburg Technical Service (Leon Dohmen, BS) for building and maintaining the equipment.
- Received May 4, 1994.
- Revision received May 17, 1995.
- Accepted June 8, 1995.
- Copyright © 1995 by American Heart Association
Hoffman BF, Rosen MR. Cellular mechanisms for cardiac arrhythmias. Circ Res.. 1981;49:1-15.
Cranefield PF, Aronson R. Cardiac Arrhythmias: The Role of Triggered Activity and Other Mechanisms. Mt Kisco, NY: Futura Publishing Co; 1988.
Kass RS, Lederer WJ, Tsien RW, Weingart R. Role of calcium ions in transient inward currents and aftercontractions induced by strophantidin in cardiac Purkinje fibers. J Physiol.. 1978;28:187-208.
Rosen MR, Gelband H, Merker C, Hoffman BF. Mechanisms of digitalis toxicity: effects of ouabain on phase four of canine Purkinje fiber transmembrane potentials. Circulation. 1973;47:681-689.
Ferrier GR, Saunders JH, Mendez C. A cellular mechanism for the generation of ventricular arrhythmias by acetylstrophanthidin. Circ Res.. 1973;32:600-609.
Gorgels APM, Vos MA, Smeets JLRM, Kriek R, Brugada P, Wellens HJJ. Delayed afterdepolarizations and atrial and ventricular arrhythmias. In: Rosen MR, Janse MJ, Wit AL, eds. Cardiac Electrophysiology: A Textbook. Mt Kisco, NY: Futura Publishing Co; 1990:341-354.
Rosen MR, Reder RF. Does triggered activity have a role in the genesis of cardiac arrhythmias? Ann Intern Med.. 1981;94:794-801.
Bhandari AK, Hong RA, Rhahimtoola SH. Triggered activity as a mechanism of recurrent ventricular tachycardia. Br Heart J.. 1988;59:501-505.
Rosen MR. Is the response to programmed electrical stimulation diagnostic of mechanisms for arrhythmias? Circulation. 1986;73(suppl II):II-18-II-27.
Wellens HJJ, Brugada P, Vanagt EJDM, Ross DL, Bär FW. New studies with triggered automaticity. In: Harrison DC, ed. Cardiac Arrhythmias: A Decade of Progress. Boston, Mass: GK Hall Medical Publishers; 1981:601-610.
Lerman BB, Belardinelli L, West A, Berne RM, DiMarco JP. Adenosine sensitive ventricular tachycardia: evidence suggesting cyclic AMP–mediated triggered activity. Circulation. 1986;74:270-280.
Sung RJ, Shapiro WA, Shen EN, Morady F. Effects of verapamil on ventricular tachycardias possibly caused by reentry, automaticity, and triggered activity. J Clin Invest.. 1983;72:350-360.
Gorgels APM, Vos MA, Leunissen JDM, Dijkman B, Smeets JLRM, Wellens HJJ. Flunarizine as a specific drug to identify triggered activity based on delayed afterdepolarizations. In: Josephson ME, Wellens HJJ, eds. Tachycardias: Mechanisms and Management. Mt Kisco, NY: Futura Publishing Co; 1993:87-97.
Chen SA, Chiang CE, Yang CJ, Cheng CC, Wu TJ, Wang SP, Chiang BN, Chang MS. Sustained atrial tachycardia in adult patients: electrophysiological characteristics, pharmacological response, possible mechanisms, and effects of radiofrequency ablation. Circulation. 1994;90:1262-1278.
Priori SG, Mantica M, Schwartz PJ. Delayed afterdepolarizations elicited in vivo by left stellate ganglion stimulation. Circulation. 1988;78:178-185.
Furukawa T, Kimura S, Castellanos A, Bassett AL, Myerburg RJ. In vivo induction of focal triggered ventricular arrhythmias and responses to overdrive pacing in the canine heart. Circulation. 1990;82:549-559.
Rosen MR, Danilo P. Effects of tetrodotoxin, lidocaine, verapamil, and AHR-2666 on ouabain induced delayed afterdepolarizations in canine Purkinje fibers. Circ Res.. 1980;46:117-124.
Felzen B, Lotan R, Binah O. Inter-species variations in myocardial responsiveness to cardiac glycosides: possible relations to the thyroid status. J Mol Cell Cardiol.. 1989;21:165-174.
Steiner CH, Kovalic ATHW. A simple technique for production of chronic complete heart block in dogs. J Appl Physiol.. 1968;25:631-632.
Vos MA, Gorgels APM, Leunissen HDM, van Deursen RTAM, Wellens HJJ. Significance of the number of stimuli to initiate ouabain-induced arrhythmias in the intact heart. Circ Res.. 1991;68:38-44.
Rosen MR. Delayed afterdepolarizations induced by digitalis. In: Rosen MR, Janse MJ, Wit AL, eds. Cardiac Electrophysiology: A Textbook. Mt Kisco, NY: Futura Publishing Co; 1990:271-281.
Berlin JR, Cannell MB, Lederer WJ. Cellular origins of the transient inward current in cardiac myocytes: role of fluctuations and waves of elevated intracellular calcium. Circ Res.. 1989;65:115-126.
Vos MA, Gorgels APM, Leunissen HDM, Wellens HJJ. Flunarizine allows differentiation between mechanisms of arrhythmias in the intact heart. Circulation. 1990;81:343-349.
Le Marec H, Spinelli W, Rosen MR. The effects of doxorubicin on ventricular tachycardia. Circulation. 1986;74:881-889.
Vos MA, Gorgels APM, Leunissen HDM, Korterink W, Wellens HJJ. Discriminative power of drugs to identify specific arrhythmogenic mechanisms in vivo. Cardiovasc Drug Ther.. 1991;5:402. Abstract.
Vos MA, Gorgels APM, De Wit B, Drenth JPH, Van Deursen RTAM, Leunissen JDM, Wellens HJJ. Premature escape beats: a model for triggered activity in the intact heart? Circulation. 1990;82:213-224.
Karaguezian HS, Katzung BG. Relative inotropic and arrhythmogenic effects of five cardiac steroids in ventricular myocardium: oscillatory afterpotentials and the role of endogenous catecholamines. J Pharmacol Exp Ther.. 1981;218:348-356.
Sheu SS, Lederer WJ. Lidocaine’s negative inotropic and antiarrhythmic actions. Circ Res.. 1985;57:578-590.
Gorgels APM, De Wit B, Beekman HDM, Dassen WRM, Wellens HJJ. Effects of different modes of stimulation on the morphology of the first QRS-complex following pacing during digitalis induced ventricular tachycardia: observations in the conscious dog with chronic complete AV-block. PACE Pacing Clin Electrophysiol.. 1986;9:842-859.