Sequential pulse internal defibrillation: Is there an advantage to “switched” current pathways?

Sequential pulse internal defibrillation: Is there an advantage to “switched” current pathways? Sequential pulse internal defibrillation delivered via orthogonal current pathways has been postulated to improve defibrillation efficacy. The efficacy of twin truncated exponential sequential shocks was compared in four different defibrillation electrode configurations in six pentobarbital-anesthetized, open-chest dogs. Lead systems consisted of the conventional single current pathway spring-patch and patch-patch electrode configurations, as well as a multiple patch lead configuration, which utilized either a single (multiple patch-unswitched) or two different (multiple patch-switched) current pathways. Curves of percent successful defibrillation versus initial voltage and energy were constructed and the data were analyzed by logistic regression. The energy and initial voltage required for 50% successful defibrillation (Es0 and V,,, respectively) for each lead system were then compared. The Em for the multiple patch-unswitched and switched lead systems (4.3 f 1.5 J and 4.6 + 1.7 J, respectively) were significantly lower than for the spring-patch (9.0 ? 3.1 J; p < .005) and patch-patch (6.6 lr 1.2 J; p < -005) lead systems. In addition, the V~O for the multiple patch-unswitched lead configuration (270.6 ? 46.4 V) was significantly lower than that of all other lead systems (p < -005). Therefore lead configurations utilizing multiple patch electrodes improve defibrillation efficacy over conventional lead systems, but there is no advantage to “switched” current pathways. (AM HEART J 1989;118:717.)

Eric S. Fain, MD, Michael Stanford

and Sunnyvale,

B. Sweeney, BA,* and Michael

R. Franz, MD.

Calif.

The treatment of recurrent and refractory ventricular tachycardia and ventricular fibrillation has been greatly improved since the development of the automatic implantable cardioverter-defibrillator. In patients with these tachyarrhythmias not controlled by antiarrhythmic drug therapy, the l-year mortality due to sudden cardiac death was reduced to 1.8 7%l as compared with 26 % to 44% 2, 3 in previous studies of comparable nonresponders to pharmacologic therapy. The device currently in use in over 4000 patients (Cardiac Pacemakers, Inc., St. Paul, Minn.) delivers a single shock via an internal defibrillation lead system pair consisting of either a helical spring electrode lying in the superior vena cava/right atrium and a left ventricular epicardial patch, or two ventricular epicardial patch electrodes. Recently, improved internal defibrillation efficacy has been reported with the use of sequential pulses

delivered via three electrodes over two different current pathways. 4-6 However, the sequential pulse system in these studies, which all utilized a right-sided intravascular defibrillation catheter and a left ventricular epicardial patch electrode, was only compared with single pulse defibrillation using the catheter alone. No comparisons with the lead systems currently in use or with other multiple electrode configurations were performed. Accordingly, we compared the efficacy of twin sequential pulse defibrillation delivered via the conventional single current pathway (spring-patch and patch-patch) electrode configurations with a single current pathway multiple patch lead system. In addition, we determined the effect of spatial separation of sequential shocks by comparing the efficacy of a multiple patch lead configuration utilizing either a single (unswitched) or two different (switched) current pathways. METHODS

From the Division of Cardiology, *Ventritex Inc., Sunnyvale. This research vale, Calif. Received Reprint Stanford 94305. 4/l/14646

was supported

for publication

Stanford

University

in part by a grant March

11, 1989;

requests: Michael R. Franz, MD, University School of Medicine,

from

accepted

Medical Ventritex June

Center;

and

Inc., Sunny-

2, 1989.

Cardiology Division-CVRC 300 Pasteur Dr., Stanford,

293, CA

Animal preparation. The experiments were conducted with six mongrel dogs of either sex with a mean weight of 26.1 + 4.7 kg (range 22.0 to 33.6 kg). The dogs were anesthetized with intravenous sodium pentobarbital 20 to 25 mg/kg, followed by 2 to 3 mg/kgihr as necessary.7 The animals were intubated and ventilated with humidified room air by a Harvard model 613 respirator (Harvard Apparatus Inc., S. Natick, Mass.) with adjustment based on arterial 717

7 18

Fain, Sweeney, and Franz

SPRING-PATCH

American

-

PATCH-PATCH

MULTIPLE

_

PATCH

Fig. 1. Defibrillation electrode configurations. A, Springpatch: right atria1 intravascular spring (anode) paired with a 7 cm2 left ventricular epicardial patch (cathode). 6, Patch-patch: shocks delivered between two 7 cm2 epicardial patch electrodes on the right (anode) and left (cathode) ventricles. C, Multiple patch: 3.5 cm2 right atrial, 3.5 cm2 right ventricular, and 7 cm2 left ventricular epicardial electrodes. “Unswitched” configuration delivered the shocks along a single current pathway between the right atria1 and right ventricular patches (common anode) and the left ventricular patch (cathode). “Switched” configuration utilized two current pathways, the first shock being delivered between the right atria1 patch (anode) and left ventricular patch (cathode) and the second shock being delivered between the right (anode) and left (cathode) ventricular electrodes. blood gases (Corning 165 pH/blood gas analyzer, Corning Glass Works, Corning, N.Y.) measured every 30 minutes to maintain oxygen tension greater than 85 mm Hg and pH between 7.35 and 7.45. Normal saline was infused at 2 ml/ kg/hr via a femoral vein throughout the experiment and normothermia was maintained with a heating blanket. A polyethylene catheter was positioned in a femoral artery and was connected to a Statham fluid-filled pressure transducer (Spectramed Inc., Critical Care Division, Oxnard, Calif.). Surface electrocardiographic leads I, II, and aVF, along with femoral arterial blood pressure, were continuously monitored and displayed on a Beckman Instruments oscilloscope (Beckman Instruments Inc., Fullerton, Calif.). A median sternotomy was performed and the heart was suspended in a pericardial cradle. A titanium spring electrode (Cardiac Pacemakers, Inc., St. Paul, Minn.) with approximately 7 cm2 surface area was inserted into the right atrium via the right atria1 appendage and was secured with a purse-string suture. A titanium circular mesh patch electrode with 7 cm2 surface area was sutured directly to the epicardial surface of the lateral left ventricle; an identical patch electrode was also sutured on the lateral right ventricle. An additional 3.5 cm2 titanium circular mesh patch

October 1989 Heart Journal

was sutured to the epicardium at the right atrial-right ventricular junction (this electrode will be referred to as the “right atrial” patch to distinguish it from the right ventricular patch mentioned above). A bipolar stimulating/ fibrillating electrode was sutured onto the right, ventricular epicardium at the outflow tract. After implantation of electrodes, the retractors were relaxed, allowing the chest to close. Defibrillation electrode configurations. Four different electrode configurations were compared and are depicted in Fig. 1, A to C. All lead systems had equal total anodal and total cathodal surface areas of 7 cm2, For all electrode configurations, each defibrillation attempt consisted of identical, sequential monophasic truncated exponential shocks of 3 msec duration separated by 0.3 msec. The “springpatch” lead system utilized the right atria1 spring as the anode and the left ventricular 7 cm2 patch as the cathode for delivery of both pulses of the sequential shock. The “patch-patch” lead system consisted of the lateral right ventricular 7 cm2 patch as the anode and the left ventricular 7 cm2 patch as the cathode for delivery of the two pulses of the sequential shock. Both the spring-patch and patch-patch lead configurations are currently used with the automatic implantable cardioverter-defibrillator (Cardiac Pacemakers, Inc.) in man. In the two other lead systems the effective surface area of the lateral right ventricular 7 cm” patch electrode was reduced to 3.5 cm2 by placing a ringshaped acetone shield (Fig. 2) beneath the patch between its epicardial fixation sutures in order to keep total anodal surface area constant at 7 cm’. The “multiple patchunswitched” lead configuration utilized the right atria1 3.5 cm2 patch tied with the right ventricular (shielded) 3.5 cm? patch as the common anode and the left ventricular 7 cm” patch as the cathode for delivery of both pulses of the sequential shock. These first three lead systems therefore used a single current pathway to deliver the two monophasic pulses of the sequential shock. Finally, the “multiple patch-switched” lead system consisted of the right atria1 3.5 cm2 patch as the anode and the left ventricular 7 cm? patch as the cathode for the first 3 msec pulse, then “switched” to deliver the second identical 3 msec pulse via the right ventricular (shielded) 3.5 cm2 patch as the anode and the left ventricular 7 cm’ patch as the cathode; again, total anodal surface area equaled 7 cm2. This final lead system therefore used two different current pathways to deliver the sequential pulses. Fibrillation/defibrillation trials. Ventricular fibrillation was induced by a l- to Z-second train of 10 msec cycle length paced current at 4 to 7 times late diastolic threshold intensity via the right ventricular epicardial stimulating electrode. The internal defibrillation lead systems were connected to a battery-operated dual-channel external cardioverter-defibrillator (Ventritex, Inc., Sunnyvale, Calif.) that could be set to deliver an initial voltage variable from 50 to 990 V in 10 V increments, with pulse duration variable from 1.~ to 20.0 msec in 0.1 msec increments; the delay between the two channel outputs can be set from 0.3 to 999 msec. The defibrillation shock consisted of two identical truncated exponential waveform pulses, each 3

Volume

118

Number

4

Sequential

pulse internal

defibrillation

7 19

Fig. 2. Epicardial patch electrodes:7 cm2(upper left) and 3.5 cm2(upper right) surfacearea.Ring-shaped acetone shield (below), when placed under the 7 cm2patch, reduces its effective surface area to 3.5 cm2.

O------• x--x A-.-A e---e

Spring-Patch Patch-Patch MuIt. Patch-Switched Mult. Patch-Unswitched

50 INITIAL

VOLTAGE

(vok)

Fig. 3. Percent successfuldefibrillation versusinitial voltage for all lead systemsin a representative animal. The raw data (open and closed circles, triangles, x’s) are shown along with the fitted curves generated by logistic regressionanalysis.

msec in duration, separated by 0.3 msec and with equal initial voltage. Only initial voltage was varied during the course of the experiment. The defibrillator measuredand displayed delivered energy and impedancefor each pulse. If the initial shock wasunsuccessful,a “rescue shock,” believed to be of sufficient energy to achieve 100% successful defibrillation, followed in lessthan 2 seconds.Only the initial shocks were used for analysis. Fibrillation/defibrillation trials were performed at 3-minute intervals, and the selected initial voltage was delivered after 15 secondsof ventricular fibrillation. The trials were recordedon a Gould ES 1000 Electrostatic recorder (Spectramed Inc., Critical Care Division) at paper speedsof 25 to 50 mm/set.

Five initial voltages in 30 to 60 V increments (range 190 to 510V) were chosenbasedon estimated transmyocardial impedance and our past experience.s These five initial voltages were tested in random order utilizing one of the four lead systems,alsochosenat random by referring to a standard table of random numbers.gEach of theseseriesof five shocks (in newly randomized order) was tested five times for eachof the four lead systemsin balancedrandom order for a total of 100fibrillation/defibrillation trials performed over the 300-minute experiment. Statistical analysis. Curves relating energy and initial voltage to percent successfuldefibrillation for each of the four lead systemsin eachdog were generatedby computer

720

Fain, Sweeney, and Franz

American

October 1989 Heart Journal

D v50 0 V80 500 -

g

400 -

e g

300 -

5 8

200 -

oLEAD

CONFIGURATION

Fig. 4. Mean initial voltages associatedwith 50%, (Vss) and 80% (V,) successfuldefibrillation for each lead configuration. *Significantly lower than all other lead systems;tsignificantly lower than spring-patch and patch-patch lead configurations. See text for level of significance.

using a logistic regressionmodello to fit the raw data to a sigmoidal dose-responsecurve, as previously described.8 The resultant defibrillation curves gave the predicted energy or initial voltage associatedwith any probability of successfuldefibrillation from 0.01 to 0.99. Differences between lead systemsin the likelihood of successfuldefibrillation were analyzed by multivariate analysis of variance for repeated measurements,followed by paired t tests to assessdifferences between individual lead systems by comparingthe meanpredicted energiesand initial voltages associatedwith 50% success(Em and Vsc,respectively) and 80% success(Esc and Vm, respectively). Changesin defibrillation successand impedance with respect to time over the courseof the experiments were analyzed by multiple logistic regression.Resultsare reported asmean k 1 standard deviation (SD) and p < 0.05 was considered indicative of statistically significant differencesin all analyses. RESULTS Defibrillation

curve determination. For each of the four electrode configurations in all animals, the set of randomly applied defibrillation trials produced a range of increasing initial voltages with their resultant energies that were generally associated with increasing percentages of success. This permitted the construction of a set of defibrillation curves for each dog relating the percent of successful defibrillations to both initial voltage and delivered energy by using logistic regression analysis to fit the raw data to “dose-response” curves. In all cases, chi square analysis demonstrated highly significant fitting of the curves to the raw data. Fig. 3 shows these curves generated by logistic regression in a representative animal.

Defibrillation curves: Initial voltage. The mean initial voltages associated with 50% (V& and 80% (Vso) successful defibrillation for each of the four lead configurations are shown in Fig. 4. Multivariate analysis of variances for repeated measurements detected significant differences between these mean values, and paired t tests were performed. The mean Vss for the multiple patch-unswitched lead configuration (270.6 k 48.4 V) was significantly lower than for the spring-patch (387.8 + 70.9 V; p < O.OOl), patch-patch (359.7 f 38.5 V;p < 0.005) and multiple patch-switched (318.0 + 60.7 V; p < 0.005) electrode systems. The multiple patch-switched lead system’s mean Vso was significantly lower than the springpatch 0, < 0.001) and patch-patch (p = 0.05) configurations, but there was no significant difference between the spring-patch and patch-patch systems. Similarly, the mean Vso for the multiple patchunswitched electrode system (305.5 + 63.9 V) was significantly lower than the spring-patch (433.0 k 86.7 V, p < O.OOl), patch-patch (429.5 + 63.8 V, p < 0.005), and multiple patch-switched (349.0 -+ 79.1 V, p < 0.05) configurations, The multiple patch-switched electrode system’s mean Vso was also lower than the spring-patch (p < 0.005) and patch-patch (p < 0.05) configurations, while there was no difference between the spring-patch and patch-patch systems. Defibrillation curves: Delivered energy. The mean energies associated with 50% (E5e) and 80% (Ego) successful defibrillation for the four lead configurations are shown in Fig. 5. Multivariate analysis of variance for repeated measurements detected signif-

Volume Number

118 4

Sequential pulse internal defibrillation

721

-r

n

1 0

)-

I-

,-

,-

I-

PATCHPATCH

LEAD

MULTIPLE PATCH“NSWlTCHED

E50 E80

MLILTIPLE Pm.xSWITCHED

CONFIGURATION

Fig. 5. Mean delivery energy (joules) associatedwith 50% (E& and 80% (Em) successfuldefibrillation for each lead configuration. *Significantly lower than spring-patch and patch-patch lead systems; tsignificantly lower than spring-patch lead configuration. See text for levels of significance.

icant differences between these mean energies, and paired t tests were performed. The mean Es0 for both the multiple patch-unswithced (4.34 & 1.48 joules) and multiple patch-switched (4.64 t- 1.69 joules) leads configurations were significantly lower than for the spring-patch (8.96 + 3.11 joules; p < 0.001 and p < 0.005, respectively) and patch-patch (6.63 +- 1.17 joules; p < 0.001 and p < 0.005, respectively) electrode systems, but the two multiple patch configurations did not differ significantly from each other. In addition, the patch-patch system’s mean Es0 was significantly lower than for the spring-patch configuration p < 0.05). Similarly, the mean Esc for both the multiple patch-unswitched (5.58 & 2.30 joules) and multiple patch-switched (5.55 * 2.21 joules) configurations were not significantly different, but were lower than for the spring-patch (10.86 f 3.91 joules; p < 0.001 and p < 0.005, respectively) and patch-patch (9.07 & 2.45 joules; p < 0.001 and p < 0.01, respectively) electrode systems. The mean Esc for the conventional spring-patch and patchpatch electrode systems were not significantly different. Impedance. For all lead systems measured, impedance varied inversely with initial voltage; increasing initial voltage resulted in decreasing transmyocardial impedance. Differences between the four lead systems were analyzed by comparing the mean impedances attained using an intermediate initial voltage of 370 V, which was used in all lead systems in all

dogs. The spring-patch (86.5 & 12.8 Q) and multiple patch-unswitched (85.3 f 7.8 0) electrode configurations had the lowest impedance, which was significantly less than the mean impedance of the patchpatch lead system (100.4 * 9.2 Q; p < 0.001 and p < 0.001, respectively). The multiple patchswitched electrode configuration had mean impedances of 121.5 f 11.3 Q for the first pulse (right atrium-left ventricle) and 110.1 + 10.1 Q for the second (right ventricle-left ventricle), which were both significantly higher than those of all other lead configurations (p < 0.001 andp < 0.005, respectively). Multiple logistic regression analysis detected a significant decline in impedance over time with all lead systems (mean 12.3 f 6.3 % for all lead systems at an initial voltage of 370 V), but no significant change in defibrillation success with respect to time. DISCUSSION

As previously shown8 defibrillation efficacy is best described as a dose-response curve rather than a distinct “defibrillation threshold.” Using these methods, we have shown that twin sequential pulses delivered via lead systems that utilized three patches have significantly lower initial voltage and energy requirements than the conventional spring-patch and patch-patch electrode configurations. In addition, we have demonstrated that there is no additional advantage to using two different current pathways to deliver the sequential shocks, as the energy require-

722

Fain, Sweeney, and Franz

American

October 1999 Heart Journal

way. If a single 6 msec pulse had been used for these other lead systems, the total delivered energy would have been reduced for any given initial voltage compared to twin 3 msec pulses (Fig. 6). Also, the “effective waveforms” would not have been identical, as the sequential shocks would have had two leading edge peak voltages and currents compared to one for the single shock. Previous

TIME

(msec) Fig. 6. Schematic diagram comparing a single 6 msec (top) shock with twin 3 msec pulses (bottom). Hatched area in the second3 msecpulserepresentsqualitatively the amount of additional energy that is delivered when twin

pulsesare administered compared to a single shock with identical initial voltage and equal total pulseduration. For example, an initial voltage setting of 320 V would result in a 35.5% increasein delivered energy when twin pulsesare utilized.

ments for successful defibrillation were comparable between the two multiple patch lead systems, while the “switched” configuration required a significantly greater initial voltage. Experimental design. The use of multiple shocks to improve defibrillation efficacy was first proposed by Wiggers in 1940,ll who delivered three to seven pulses 1 to 2 seconds apart via internal cardiac paddles. In the present study, we used twin pulses separated by 0.3 msec, as it has been shown that defibrillation efficacy is best when pulse separation is less than or equal to 1 msec.4, l2 The experiment was designed to evaluate and compare the efficacy of the different lead configurations themselves while holding all other variables constant. All four lead configurations were utilized in each dog, permitting the construction of a set of four defibrillation curves, and allowing each animal to serve as its own control. Total anodal and total cathodal electrode surface areas were made identical (7 cm2) in all lead systems, as increased surface area has been shown to decrease defibrillation energy requirements.131 l4 In all lead systems, the waveforms were identical and total pulse duration was held constant at 6 msec. In addition, all lead configurations used twin sequential shocks, even though all but the multiple patch-switched system delivered both shocks over the same current path-

sequential

pulse defibrillation

investigations.

Previous investigations of the efficacy of sequential defibrillation using implantable lead systems have been performed by Jones et a1.,4a5, I2 Kallok et al.,l’$T ” Bourland et al.,” and Zipes et a1.16 All compared a single shock delivered via an intravascular catheter (Medtronic 6880, Minneapolis, Minn.) positioned with its cathode at the right ventricular apex and its anode at the superior vena cava-right atria1 junction with twin sequential shocks using two different current pathways; the first pulse was delivered via the catheter and the second was applied between the catheter tip at the right ventricular apex and a left ventricular epicardial patch electrode. The sequential pulse technique was found to be superior to the catheter in all studies. However, no comparisons to the conventional spring-patch or patch-patch lead systems or to multiple electrode, single current pathway configurations were made. Catheter defibrillation is thought to be less effective because current is shunted through the low impedance blood pool away from the higher impedance myocardium.” In fact, a single defibrillation shock between the catheter tip and left ventricular epicardial electrode (the second pulse of the sequential shock) has been shown4,17’18 to have lower energy requirements than catheter defibrillation, and one study18 found catheter defibrillation to be the least efficacious of seven defibrillation techniques. In addition, a comparison of data from two studies demonstrated a significant reduction in energy require-

ments using the spring-patch lead system when compared with catheter defibrillation.lg Therefore catheter defibrillation is not optimal, and the conclusion that sequential pulse defibrillation over two different current pathways improves defibrillation efficacy should not be based on this comparison.

Recently, Chang et al.18 compared 13 different lead configurations composed of catheter, epicardial, and subcutaneous electrodes, which utilized either single or sequential shocks. Sequential pulse defibrillation with a lead system identical to that used previously46.12,14-16and as described above, as well as a lead configuration utilizing multiple epicardial patches and two current pathways, were found to have superior defibrillation efficacy compared with all other

Volume Number

118 4

Sequential

lead systems. However, the multiple patch system was not compared with an identical configuration that used a single current pathway. In addition, their study and those cited above failed to control for many confounding variables including total anodal and cathodal electrode surface area, “effective waveform” and shock polarity, while only one study controlled for total pulse duration.6 Theoretical

basis of sequential

pulse defibrillation.

It

has been hypothesized4-6 that improved efficacy with sequential defibrillation shocks delivered over “orthogonal” current pathways is due to both temporal and spatial separation of the pulses, resulting in a more uniform current density and more widespread current distribution. With the use of a multiple patch electrode system, we have shown that spatial separation of the defibrillation pulses produced no additional advantage in terms of energy requirements, while it required significantly greater initial voltages. The higher initial voltages needed for successful defibrillation were probably due for the most part to the higher impedance of the “switched” lead system. Therefore utilization of an optimal single current patchway, which involves both right and left ventricular free walls as well as the interventricular septum, can produce high effective defibrillation. The multiple patches provide an advantage over the conventional patch-patch lead configuration, most likely by providing a more uniform current density across both ventricles. In fact, it has been hypothesized that defibrillation efficacy should be improved if a uniform and sufficient potential gradient is generated throughout the ventricles, regardless of electrode configuration.20 In addition, the decrease in impedance without change in electrode surface area results in an increase in delivered energy. It is not known if there is an advantage to temporal separation of sequential pulses over a single current pathwaythat is, using twin 3 msec pulses versus a single 6 msec shock with the initial voltage increased to deliver an equal amount of energy. Early experiments by Kugelbergzl and Resnekov et a1.22suggested that energy requirements may be decreased with the use of twin pulses over the same pathway. On the other hand, Geddes et a1.,23 Schuder et a1.,24 and Moore et a1.25 demonstrated either no reduction or higher energy requirements using similar configurations. In addition, Chang et al. la showed no significant differences in defibrillation energy requirements when either single or twin truncated exponential shocks were delivered via a right-sided intravascular catheter. Although no formal comparison was performed during this investigation, our experience is that there is no significant difference in efficacy between twin

puLse internal

defibrillation

723

pulse and single pulse defibrillation administered over a single current pathway when total shock duration and total delivered energy are equal. Clinical applications. The results of this study have important clinical applications. When the multiple patch-unswitched lead configuration is used, defibrillation energy requirements are greatly reduced. In addition, because of the low impedance of the system, a lower initial voltage can be used to deliver the needed energy. Currently, the size of the pulse generator of the automatic implantable defibrillator is limited by the size of the batteries and energy storage capacitors. The lower initial voltage requirements would allow a reduction in size of both of these elements. If a single pulse rather than sequential shocks is utilized, the initial voltage would have to be increased to deliver the same energy, but a set of capacitors could be eliminated. Therefore the efficacy of the device can be improved and/or the size of the pulse generator decreased by the use of this lead system. In addition, by reducing the shock’s initial voltage and current, myocardial damage can also be diminished.26s 27 Study limitations. A potential limitation of our study is that only healthy dogs with normal myocardium were used. However, previous studies that used different defibrillation lead configurations did not demonstrate a significant effect of acute ischemia or subacute myocardial infarction on defibrillation success.i7, la Also, these experiments were performed for a short period. It is not known if the results would have been different if the leads were implanted for a long term, as defibrillation energy requirements have been shown to change with time.lg In addition, since only one sequential shock interpulse separation was evaluated, it is not known if the results would have been different if other pulse separations were tested. Conclusions. We have demonstrated in dogs that lead configurations that use multiple patch electrodes to deliver twin sequential pulses have significantly lower initial voltage and energy requirements than conventional lead systems. However, employing different current pathways to deliver the sequential shocks provided no additional advantage and required higher initial voltages for successful defibrillation. Consequently, if defibrillation in man behaves similarly, the efficacy of an automatic implantable defibrillator may be improved and the size of the device may be reduced. We thank Roger A. strong support, Robert nical assistance, David and Erin Howard for

Winkle, MD, and Kathi Senelly for their Kernoff and George Snidow for their techAhn, PhD, for his computer programming, preparing the manuscript.

October

724

Fain, Sweeney, and Franz

American

REFERENCES

1. Echt DE, Armstrong K, Schmidt P, Oyer RE, Stinson EB, Winkle RA. Clinical experience, complications, and survival in 70 patients with the automatic implantable cardioverter: defibrillator. Circulation 1985;71:289-96. 2. Swerdlow CD, Winkle RA, Mason JW. Determinants of survival in patients with ventricular tachyarrhythmias. N Engl .J Med 1983;308:1436-42. TB, Lown B, Podrid PJ, DeSilva R. Long-term sur3. Graboys vival of patients with malignant ventricular arrhythmias treated with antiarrhythmic drugs. Am J Cardiol1982;50:43743. 4. Jones DL, Klein GJ, Kallok MJ. Improved internal defibrillation with twin pulse sequential energy delivery to different lead orientations-in pigs.Am J Card&l 1985;551821-5. 5. Jones DL. Klein GJ. Guiraudon GM. et al. Internal cardiac defibrilla$on in man; pronounced improvement with sequential pulse delivery to two different lead orientations. Circulation-1986;73:484-91. 6. Bourland JD. Tacker WA. Wessale JL. Kallok MJ. Graf JE. Geddes LA. Sequential pulse defibrillation for implantable defibrillators. Med Instrum 1986;20:138-42. I. Babbs CF. Effect of pentobarbital anesthesia on ventricular defibrillation threshold in dogs. AM HEART J 1978;95:331-7. 8. Davv JM. Fain ES Dorian P. Winkle RA. The relationshin between successful defibrillation and delivered energy in open-chest dogs: reappraisal of the “defibrillation threshold” concept. AM HEART J 1987;113:77-84. 9. Snedecor GW, Cochran WG, editors. Statistical methods. 7th ed. Ames, Iowa: The Iowa State University Press, 1980:463-6. 10. Reinhardt PS, editor. User’s guide. SAS supplemental library. Cary, North Carolina: SAS Institute, Inc., 1980:83-102. 11. Wiggers CJ. The physiologic basis for cardiac resuscitation from ventricular fibrillation-method for serial defibrillation. AM HEART J 1940;20:413-22. JD, Tacker WA, Kallok MJ, 12. Jones DL, Sohla A, Bourland Klein GJ. Internal ventricular defibrillation with sequential pulse countershock in pigs: comparison with single pulses and effects of pulse separation. PACE 1987;10:497-502. 13. Mead RH, Echt DS, Stinson EB, Schmidt P, Winkle RA. The automatic implantable defibrillator: improved defibrillation and lower impedance using two large patch leads [Abstract]. J Am Co11 Cardiol 1985;5:455. 14. Kallok MJ, Bourland JD, Tacker WA, Jones DL, Klein GJ. Wessale JL. Optimization of epicardial electrode size and im-

15.

16.

17. 18.

19.

20.

21. 22.

23.

24.

25.

26.

27.

Heart

1989 .lournal

plant site for reduced sequential pulse defibrillation thresh bids. Med Instrum 7986;20:36-9. Kallok M-J. Tacker WA. ,Jones DL. Klein GJ. Wessale Jr,. Transvenous electrode systems and sequential pulse stimulation for reduced implantable defibrillator thresholds [Al) stract]. J Am Co11 Cardiol 1985;5:456. Zipes DP, Kallok MJ, Gill RM, Kammerling JM. Efficacy of sequential electrical shocks to terminate ventricular fibrillation in dogs after myocardial infarction [Abstract]. Circulation 1984;7O(Suppl IIl:II-407. Jones DL, Yohla A, Klein (;J. Internal cardiac defibrillation threshold: effects of acute ischemia. PACE 1986:9:322-31. Chang MS, Inove H, Kallok MJ, Zipes DP. Doubie and triple sequential shocks reduce ventricular defibrillation threshold in dogs with and without myocardial infarction. .J Am Co11 Cardiol 1986;8:1393-405. Fain ES, Billingham M, Winkle RA. Internal defibrillation: histopathology and temporal stability of defibrillation energy requirements .I Am Co11 Cardiol 1987;9:631-8. Wharton ?JM. Wolf PD. Chen PS. et al. Is an absolute minimum potential gradient required for ventricular defibrillation? [Abstract] Circulation 1986;74(Suppl II):II-342. Kugelberg J. Ventricular defibrillation-a new aspect. Acta Chir Stand 1967;372(Suppll:l-93. Resnekov I,, Norman J, Lord I’. Sowton E. Ventricular defibrillation by monophasic trapezoidal-shaped doublepulses of low electrical energy. Cardiovasc Res I968;3:261-4. Geddes L.4. Tacker WA. McFarlane JR. Ventricular defibrillation with single and twin pulses of half-sinusoidal current. J Appl Physiol 1973;34:8-11. Schuder JC, Stoeckle H, Prabhakar KY, Gold dH. Chier MT, West JA. Transthoracic ventricular defibrillation in the dog with unidirectional rectangular double pulses. Cardiovasc Res 1970;4:497-503. Moore TW, Di Meo FN, Dubin SE. The effect of shock separation time on multiple-shock defibrillation. Med Instrum 1978;12:31-3. Doherty PW, McLaughlin PR, Billingham M, Kernoff R, Goris ML, Harrison DC. Cardiac damage produced by direct current countershock applied to the heart. Am .J Cardiol1979;43:22532. Barker-Voelz MA, Van Vleet JF, Tacker WA, Bourland JD. Alterations induced by a single defibrillating shock applied through a chronically implanted catheter electrode. d Electrocardiol 1983;16:167-80.