Termination of ventricular fibrillation in dogs by depolarizing a critical amount of myocardium

EXPERIMENTAL STUDIES

Termination of Ventricular Fibrillation in Dogs by Depolarizing a Critical Amount of Myocardium

DOUGLAS P. ZIPES, MD, FACC JOHN FISCHER, MD ROBERT M. KING, MD ANN deB. NICOLL, RN WALTER W. JOLLY, MD Indianapolis, Indiana

From the Krannert instituteof Cardiology,Marion County General Hospital, the Department of Medicine and the Department of Surgery, Indiana University School of Medicine and the Veterans Administration Hospital, Indiinapolis, Ind. This study was supported in part by the Herman C. Krannert Fund, Indianapolis, Ind., Grants HL06306. HL-05663 and HL-05749 from the National Heart and Lung Institute, National Institutes of Health, Bethesda, Md. and the Indiana Heart Association, Indianapolis. Ind. Manuscript accepted September -19, 1974. Address for reprints: Douglas P. Zipes, MD, Indiana University School of Medicine, 1100 West Michigan St., Indianapolis, Ind. 46202.

The role of a critical myocardial mass required to maintain ventrkular fibrillation initiated by rapid ventricular pacing was studied by two methods in dogs placed on total cardiopulmonary bypass. In the first method, depolarization of a limited myocardlal mass was accomplished by injecting potassium chloride into one or two coronary arteries. Injection of potassium chloride simultaneously into the left circumflex and left anterior descending coronary arteries abollshed ventrlcuiar fibrillation more often than did injectlon into any other single or combination of two coronary arteries (P SO.0001). Ventricular flbrillation could not be reinitiated as long as the lefl ventricle remained inexcitable. Immersing the heart in a solution of potassium chloride or injecting the solution into the right and left ventricular cavities failed to terminate ventricular fibrillation. The second method evaluated the amount of current necessary to terminate ventricular fibrillation when the current was passed between two right ventricular electrodes, between two lefl ventricular electrodes and between one right ventricular and one lefl ventricular electrode. Electrical shocks of equal magnttude terminated ventricular fibrillation most often when those shocks were delivered between an electrode located at the right ventricular apex and an electrode located at the posterior base of the lefl ventricle, and least often when the shock was delivered between two right ventricular electrodes:Successful defibrillation results when a critical amount of myocardium becomes depolarized by etther potassium chloride or electrical discharge; depolarization of every cell in both ventricles is not necessary to terminate ventricular fibrillation in the entlre heart.

Although the explanation for the pathogenesis of self-sustained fibrillation remains divided between the ectopic focus and reentry theories,‘e3 much evidence exists to substantiate the ‘need for a critical myocardial mass to maintain self-sustained fibrillation. A critical mass represents the amount of myocardium that must be present and excitable for fibrillation to continue spontaneously. In the late 1800’s, McWilliam4 and Porter5 noted that ventricular fibrillation was more easily supported in larger hearts and, in 1914, GarreyG demonstrated that when pieces of ventricular muscle with a surface area of less than 4 cm2 were shaved from the wall of the fibrillating left ventricle, these pieces stopped fibrillating. The remaining portion of the ventricle continued to fibrillate until three fourths of the ventricular mass had been removed. In other experiments, after appropriate cuts divided the fibrillating heart into the right ventricle, interventricular septum and left ventricle, the right ventricle stopped fibrillating while fibrillation in the larger masses continued. The purpose of this study was to test whether ventricular fibrillation can be terminated in the intact heart in situ by depolarizing only a limited population of cells. If ventricular fibrillation can be termi-

July 1975

The American Journal of CARDIOLOGY

Volume 36

37

VENTRICULAR CEFlBRlLLATlON-ZII’ES ET AL.

nated in such a fashion, it indicates that the remaining excitable cells represent a critical mass that is insufficient to maintain self-supported ventricular fibrillation. Such a concept has been suggested to explain observations recorded during defibrillation studies in dogs using an intracavitary catheter electrode.7 The results of our study provide direct proof in support of this concept. Two types of experiments were devised. In one group of dogs we employed a solution of potassium chloride to depolarize a limited mass of myocardium by bolus injection of the solution into the right, left anterior descending or circumflex coronary artery. In the second group of dogs, we determined the amount of current and voltage necessary to terminate ventricular fibrillation in the entire heart when the current was passed in the form of a trapezoidal wave between two electrodes sewn on the right ventricle, between two electrodes sewn on the left ventricle and between various combinations of one right ventricular and one left ventricular electrode.

Methods Healthy mongrel dogs of either sex weighing 15 to 30 kg were anesthetized with secobarbital (30 mg/kg body weight), intubated and ventilated with a Harvard respirator at a rate and tidal volume predicted from a nomogram. The chest was opened in the midline, and the pericardium was incised and retracted to cradle the heart. Total cardiopulmonary bypass was achieved in both groups of animals. After ligation of the azygous vein and administration of heparin (3 mg/kg intravenously), large bore catheters to collect venous blood were inserted into the superior and inferior venae cavae through a right atriotomy and secured by purse-string sutures. Umbilical tapes, placed loosely around the venae cavae, were later tightened to direct all venous blood returning in the venae cavae into the venous catheters. Blood was also collected from a suction catheter placed in the chest cavity and from a left ventricular vent catheter and pumped (Sigma motor model T65, Leland Electric) into a filtered cardiotomy reservoir (model Ql20, Bentley Laboratory) that emptied into the venous return line. Venous blood was drained by gravity into a disposable blood oxygenator (pediatric size Temptrol, model $110, Bentley Laboratory), primed with saline and Ringer’s lactate solution through which 100 percent oxygen was bubbled. The blood, warmed and oxygenated in the Temptrol oxygenator, was pumped by a finger pump (Sigma motor model TMl, Leland Electric) into a large arterial perfusion catheter that had been placed in the abdominal aorta by way of the femoral artery. Arterial flow rates were kept between 2 and 3 liters/min. Systemic blood pressure was monitored on an oscilloscope by means of a strain gauge transducer (Statham 23BD) that was connected to an indwelling catheter placed in the carotid artery. Mean systolic blood pressure was maintained between 45 and 65 mm Hg by adjusting venous return, altering arterial flow rate or by infusing normal saline solution; vasopressors were not used. The body temperature of the dogs was kept between 98 and 101’ F. Periodic blood gas and electrolyte determinations during the experiments revealed that partial pressure of oxygen was always greater than 100 mm Hg. Initial hypokalemia in some animals, possibly related to hyperventilation before

38

July 1975

The American Journal of CARDIOLOGY

Volume 36

institution of cardiopulmonary bypass, was gradually corrected. In the group of dogs that received intraarterial injections of potassium chloride, serum potassium concentrations slowly rose and reached levels of 5 to 7 mEq/liter by the end of the experiment. In the second group of’animals, serum potassium levels gradually rose toward normal but rarely exceeded 4 mEq/liter. Plasma levels of sodium and chloride became elevated as a result of saline infusion. Arterial pH was kept above 7.25 with periodic infusion of sodium bicarbonate. The hematocrit level fell in most dogs as a result of blood loss, pump-induced hemolysis and dilution with saline infusion. However, the experimental results we obtained cannot be ascribed to metabolic or hematologic imbalances since comparisons were made between interventions performed in the same animal within limited time periods, during which abnormal metabolic or hematologic status would be expected to affect the entire heart more or less equally. Electrical activity was recorded with bipolar J-shaped electrodes that were hooked into the epicardium of the right or left atrium, or both, and in the following epicardial ventricular sites: high right ventricle near the outflow tract, low right ventricle along the inferior border, posterobasal left ventricle, in the lateral free wall of the left ventricle adjacent to the interventricular septum, and apex of the left ventricle. The latter two left ventricular electrodes were placed to reflect distribution of the left anterior descending coronary artery, whereas the posterobasal electrode reflected distribution of the circumflex artery. Bipolar electrical activity was filtered between 40 and 500 hertz, displayed along with lead II of the electrocardiogram on a switched beam oscilloscope (DR8, Electronics for Medicine), and recorded on photographic paper at speeds of 50 and 100 mm/set. After the animal was placed on total cardiopulmonary bypass, we initiated ventricular fibrillation by rapidly stimulating either the right or left ventricle at rates of 20 to 100 times/set for 5 to 30 seconds. Stimuli, obtained from a pulse generator through an isolation transformer (Digipulser, model 830; Isopulser, model 850, WP Instrument, Inc.), were rectangular pulses of 2 msec duration delivered over bipolar electrodes at an intensity of 1 l/2 to 2 times the diastolic threshold. In the potassium studies, stimuli of 30 volts and 5 milliamperes were sometimes used. Ventricular fibrillation was considered present when totally disorganized ventricular activity continued spontaneously for a minimum of 15 to 30 seconds after cessation of ventricular stimulation. Ventricular fibrillation was considered terminated when either organized electrical activity or complete absence of all electrical activity replaced ventricular fibrillation for more than 5 seconds, simultaneously in all five of the epicardial ventricular leads and lead II. Potassium

Studies

In the initial group of eight dogs treated with potassium chloride infusion, polyethylene tubing (22 gauge, Deseret) was inserted by way of a small arteriotomy into the right coronary artery near its origin and into the midportion of the left anterior descending coronary artery. The tip of the tubing in the latter vessel was advanced in retrograde manner to the bifurcation of the main left coronary artery, where a single potassium chloride injection depolarized ventricular muscle located in areas perfused by both the circumflex and left anterior descending coronary arteries. Brief occlusion at the origin of the right coronary artery during retrograde infusion into the left anterior descending

VENTRIUJLAR DEFWWAWN-ZPES

artery prevented reflux of potassium chloride into the former vessel, so that the solution perfused primarily the left ventricle. If only the anterior or posterior left ventricular recording, but not both, was affected by the potassium chloride, the infusion was considered a single vessel infusion made into either the left anterior descending or circumflex coronary artery, respectively. In order to perfuse these vessels separately, in the final group of six animals subjected to potassium chloride infusion, the origins of the right, left anterior descending and circumflex coronary arteries were dissected free from the epicardium and each vessel was cannulated individually. The tips of the tubing, located within 1 to 2 cm of the origin of each coronary arand secured in place tery, were directed “downstream” with epicardial or purse-string sutures. The small diameter of the polyethylene tubing allowed unimpeded coronary blood flow. To determine whether depolarization of a limited myocardial mass could result in termination of ventricular fibrillation in the remaining myocardium, a fixed total amount of potassium chloride was injected into a single coronary artery, or equal portions of the total amount of potassium chloride were injected into two coronary arteries simultaneously. Thus, the distribution of the particular coronary artery or combination of coronary arteries perfused with potassium chloride determined the amount of myocardium that was depolarized. Although anastomoses between coronary arteries permitted a certain amount of cross circulation, the major portion of the ventricle not directly perfused remained unaffected by infusion of potassium chloride into the contralateral side, as judged by the epicardial recordings. In each experiment, we compared the efficacy of equal total amounts of potassium chloride in terminating fibrillation when the solution was injected into one or two coronary arteries. Finally, in another group of animals, potassium chloride (25 to 50 ml, 1 mEq/ml) was injected into the cavity of the left or right ventricle, or both, or the entire heart in situ was immersed in a polyethylene “bag” filled with 160 ml of potassium chloride (1 mEq/ml) for 30 seconds. Electrical Defibrillation Studies In the group of 11 animals electrically shocked, pairs of circular electrodes (3.14 cm2) were sewn on the right and left ventricular epicardium, in a plane parallel to the interventricular septum. The electrodes were cut from copper sheets and coated with a layer of solder; the copper did not come into direct contact with the myocardium. On the right ventricle, the electrodes were sewn close to the atria1 border near the outflow tract (RVnaSe, RVn) and near the inferior margin of the right ventricular apex (RVA~~~, RVA). On the left ventricle, they were sewn in the posterobasal area between the origins of the left anterior descending and circumflex coronary arteries (LVnase, LVn), and near the left ventricular apex (LVA~~~, LVA). The current and voltage wave forms were recorded on a calibrated storage beam oscilloscope (Tektronix model Dll) from which the amplitude was measured. All electrical grounds, including the right leg electrode for the electrocardiogram, were removed from the animal. The shock was delivered after ventricular fibrillation had been initiated and was present spontaneously for at least 30 seconds. We determined the current and voltage required to terminate ventricular fibrillation in the entire heart when the shock was delivered between the following pairs of electrode combinations: RVn-RVA, LVn-LVA, RVn-LVn, RVA-LVA, RVn-LVA and

ET AL.

70 POTASSIUM

60 -

50 RCA 8 r$

p d.04

LAD+LC

pc.cml

40-

B 8 k

30 -

ep ”

-L

20 -

__

38

II si-

10 -

-L

0

L -+-

L4!

RCA+ LAD

RCA+ LC

LAD+

LC

FIGURE 1. Graph indicating the percent of potassium chloride injections that successfully terminated ventricular fibrillation (Y axis) according to the single or combination of coronary arteries into which the solution was injected (X axis). Figures above the bars show the number of injections that terminated ventricular fibrillation (numerator) and the number of total injections (denominator). VF = ventricular fibrillation.

RVA-LVn. In addition, each electrode of each pair served as the “hot” electrode; that is, the current and voltage required to terminate fibrillation was determined separately for current flowing in each direction between two electrodes. The distance between each of the electrode pairs was measured and, at the conclusion of each experiment, the heart was removed and weighed. The electrical impulse used was a truncated exponential (“trapezoidal pulse”) impulse delivered from a capacitor (800 rfarad at 600 volts direct current) that had been charged to a selected voltage with a variable regulated power supply. The length of the pulse (24 msec) was controlled by a variable integrated circuit timer that operated a silicon-controlled rectifier circuit.* Results Potassium Studies The results from the entire group of 14 dogs were combined and are presented in Figure 1. Injections into the right coronary artery resulted in significantly fewer terminations than injections at any other site (P <0.04), whereas the 72 percent rate of termination during simultaneous injection into the left anterior descending and circumflex coronary arteries was sigThe electrical circuitry was designed and built by Medical Electronics Consulting Associates, Indianapolis, Ind. l

July 1975

The American Journal of CARMOLOGY

Volume 36

39

VENTFWJLAR DEFt9RtLLATfON-ZtPES ET AL.

FIGURE 2. Termination of ventricular fibrillation by selective potassium chloride infusion. A, potassium chloride (l/2 mEq) injected simultaneously into both the right (RCA) and left anterior descending (LAD) coronary arteries at arrow. Electrical activity recorded at the LRV, ALV and LLV electrodes became suppressed. Ventricular fibrillation continued. 6, potassium chloride injected simultaneously into both the right and circumflex coronary arteries. Electrical activity recorded at the LRV and PLV electrodes became suppressed. Ventricular fibrillation was terminated. C, potassium chloride injected simultaneously into both the left anterior and circumflex coronary arteries. Electrical activity recorded at the ALV, LLV and PLV electrodes became suppressed. Ventricular fibrillation terminated. LA = left atrium: HRV = high right ventricle near outflow tract; LRV = low right ventricle along the inferior border; ALV = apex of left ventricle reflecting distribution of left anterior descending artery; LLV = free wall of the lateral left ventricle adjacent to the interventricular septum, reflecting distribution of the left anterior descending artery; PLV = posterior left ventricle reflecting distribution of the circumflex coronary artery. II = lead II of the electrocardiogram. Time lines 1 second. Paper speed 50 mmkec.

nificantly greater than terminations produced by any other injections (P
40

July 1975

The American Journal of CARDIOLOGY

Volume 36

the left anterior descending plus the circumflex coronary artery. Therefore, the remaining number of excitable cells represented a critical mass insufficient to maintain fibrillation. Infusion of potassium chloride into only one coronary artery or simultaneously into the right and left anterior descending coronary arteries failed to terminate ventricular fibrillation, presumably because the remaining excitable mass was large enough to support continued fibrillation. In this particular example, the high right ventricle and left atrium were not affected by infusions of potassium chloride into the right and circumflex coronary artery, respectively, because these areas were supplied by branches close to the origin of each coronary artery and were not perfused by the potassium chloride solution. When ventricular fibrillation terminated after infusion of potassium chloride into the left coronary artery, the refractory period and diastolic stimulation threshold for the right ventricle remained at prefibrillation control values. However, rapid right ventricular stimulation failed to initiate ventricular fi-

brillation as long as the left ventricle remained depolarized by the effects of the potassium chloride (Fig. 3, A and B). After the return of left ventricular activity, fibrillation could be promptly restarted (Fig. 3C). Right ventricular inexcitability after infusion of potassium into the right coronary artery did not prevent precipitation of left ventricular fibrillation (not shown). Ventricular fibrillation could not be terminated when potassium chloride (25 to 50 ml, 1 mEq/ml) was injected into the cavity of the right or left ventricle and confined there for 15 to 30 seconds. Similarly, ventricular fibrillation could not be terminated by immersing the entire heart into a plastic bag with a potassium chloride solution (160 ml, 1 mEq/ml) (not shown). Electrical Defibrillation Studies The combination of electrodes, the distance between electrodes and the heart weight were all major factors affecting the ability of the electroshock to defibrillate. The percent termination of ventricular fibrillation in 1,170 shocks for three current and voltage ranges is presented in Figure 4, in which each curve represents one of the electrode combinations. The rate of termination of ventricular fibrillation is significantly different among the six electrode combinations at each range for both voltage and amperes (overall P value
ment, the distance between RVn and RVA equaled the distance between LVn and LVA (3 to 5 cm). However, the LVn-LVA electrode combination yielded a higher percent of defibrillations for electroshocks of equal magnitude than did the RVn-RVA combination (P
Discussion Critical ventricular mass required to maintain ventricular fibrillation: The data from the potassium infusion studies clearly establish for the intact dog heart in situ the importance of a critical ventricular mass to maintain self-supported ventricular fibrillation, which was initiated by rapid electrical stimulation. Ventricular fibrillation was terminated when potassium infusion reduced the remaining amount of excitable myocardium to a value less than that of the critical mass. Therefore, depolarization of every cell in both ventricles is not necessary to terminate ventricular fibrillation in the entire heart. The amount of this critical mass cannot be accurately quantified from our protocol, since we did not correlate combinations of potassium chloride injections that terminated fibrillation with the exact amount of myocardial mass perfused. This might be accomplished by injecting different dye markers into the right, left anterior descending and circumflex coronary arteries after completion of the potassium infusions. However, if we assume from Figure 1 that the entire left ventricle and interventricular septum, or a comparable amount of myocardium, must be depolarized to reduce the remaining excitable myocardium to a value insufficient to maintain ventricular fibrillation, then the critical mass necessary for ventricular fibrillation to continue in 13 dogs studied represents 28 percent of the total ventricular myocardium, a percentage remarkably similar to that found by Garrey.6 This value is at best only an estimate for the dog heart because of the problems generated by cross circulation, our inability to perfuse branches arising very close to the origin of the coronary artery, and the fact the potassium chloride very probably did not uniformly depolarize the entire ventricular mass with which it made contact. This percentage may not necessarily be extrapolated directly to the patient with heart disease because the tendency to perpetuate fibrillation in the latter situation might be affected by many variables. Support of concept of critical myocardial mass: The data from the electroshock studies provide indi-

July 1975

The American Journal of CARDIOLOGY

Volume 36

41

FIGURE 3. Termination of ventricular fibrillation by infusion of potassium chloride into the left anterior descending and circumflex arteries with faiture to initiate ventricular fibrillation white the left ventricle remained inexcitable. Pen& A and B are a continuous recording with the terminal two complexes in panel A repeated as the first two complexes in panel B. Potassium chloride (1 mEqfmf, 1 ml) was injected simultaneously km the left anterior descending and circumflex coronary arterim at arrow. Electrical activity recorded at the LLV, PLV and ALV electrodes became suppressed and an arganized ventricular rhythm replaced ventricular fibrillation in the right ventricle. B, rapid rtght ventricular pacing (3O/sec for about 4 seconds] failed to initiate ventricular fibrillation. C, after recovery of left ventricular excitability, right ventricular pacing (3Olsac for about 3 l/2 seconds) promptly initffted ventricular fibrillation. RA = right atrium; other abbreviations as in Figure 2, “Time tines 1 second. Paper speed 100 mm/see.

RELATIONSHIP VF

I

TERMINATION,

BETWEEN CURRENT

AND

VOLTAGE

B

A

80-

60504030zolo-

FIGURE 4. Percent of electroshocks that successfully terminated ventricular fibrillation (Y axis) for three voltage and current ranges (X axis), according to tie electrode combination used. A, amperes: 9, voltage. VF = ventricular fibrillation.

1 .oo1 .ss

2.002 .ss

3 .oo3.99

loo199

200299

AMPERES

300399

VOLTAGE

RVA-LVA

EP44-112073

I

:

FIGURE 5. Electroshock termination of ventricular fibrillation with electrode combinationRVALV* (right and left ventrfcular apex). Top traclngs. subthreshold shock (arrow) caused transient regularization of recorded ventricular activity but ventricular

fibrillation

continued.

Bat-

tom irachgs, threshold shock (arrow) terminated ventricular fibrillation after a brief burst of repetitive ventricular activity. Atrial fibrillation was also terminated. Paper speed 50 mm/set. Horizontal line = 200 msec. A = amperes; V = volts; ottrer abbreviations as in Figure 2.

July 1975

l’ho Amerkan

Journal of CARMOLGGY

Vdume 36

43

VENTRlCUAR DEFtBRlLLATlCN-i!lFES ET AL.

rect evidence supporting the concept of a critical myocardial mass. Those data indicate that electrical shocks of equal magnitude terminated ventricular fibrillation most often when the shocks were delivered between the RVA-LVs electrodes and least often when they were delivered between the RVn-RVA electrodes. LVn-LVA and the remaining electrode combinations resulted in intermediate values. If these observations are explained in terms of the critical myocardial mass concept, then we would postulate that current flowing between two right ventricular electrodes (RVn-RVJ encountered, and therefore depolarized, the least number of cells, and resulted in the lowest rate of cardioversions. Current traveling between the right ventricular apex and posterobasal left ventricle (RVA-LVn) encountered, and therefore depolarized, the largest number of cells and resulted in the highest rate of cardioversions. Differences between the remaining electrode combinations may have been less great in terms of mass bounded by the two electrodes and resulted in intermediate values with a fair amount of overlap. Distance between electrodes played a role in terminating fibrillation because interelectrode distance influenced the amount of mass through which the current flowed. However, the electrode combination seemed more important. The distance between RVn-RVA and LVn-LVA was equal in each experiment. Nevertheless, a higher percentage of terminations occurred with the LVn-LVA combination, presumably because of the larger mass bounded by the two left ventricular electrodes. The RVA-LVn electrode combination most successfully terminated ventricular fibrillation for two reasons. First, the combination spanned virtually the entire left ventricle and, second, its interelectrode distance was generally the largest. Clinical implications: Speculations about the events that occur during clinical direct-current electroshock are consistent with the concepts advanced

in this report. For example, during electrical defibrillation, the shock cannot discharge many cells that have just been depolarized spontaneously (by the fibrillating rhythm) and are absolutely refractory at the moment the shock is delivered. These cells naturally repolarize in advance of cells that are discharged at a slightly later time by electroshock, but, during a successful cardioversion, they represent a mass of cells insufficient to perpetuate the fibrillation. Continued fibrillation indicates failure to depolarize a sufficient number of cells. Administering a continuous series of shocks has been suggested to terminate refractory ventricular fibrillation.8vg It is possible that each successive shock depolarizes more muscle mass until finally a sufficient amount of myocardium becomes depolarized and fibrillation stops. Heart weight also has been important during defibrillated studies performed in man. The heavier the heart, the greater the energy required for defibrillation.‘O This concept of critical mass may also explain the success of intracavitary catheter defibrillation in man and animals,7,10 because depolarization of the entire heart is not necessary to terminate fibrillation. From our data, however, we would predict that energy requirements for defibrillation with a catheter electrode in or behind (coronary sinus) the left ventricle might be less than for a catheter electrode in the right ventricle. Such comparisons have not been performed by Mirowski and Mower.12 Acknowledgment We are grateful to Fred Steinhoff, Jim Craig and Louis Linthecombe of Medical Electronics Consulting Associates for help with the electrical aspects of the study, to Dan Lowe, MD for surgical assistance in some of the experiments, and to Harry M. Brittain for help with the statistical analyses. Summer students James Linderman, Teresa Plank and David Murray also worked on this project and were valuable assistants.

References 1. Wiggers CJ: The mechanism and nature of ventricular fibrillation. Am Heart J 20:399-412. 1940 2. Scherf D: Studies on auricular tachycardia caused by aconitine administration. Proc Sot Exp Biol Med 64:233-239, 1947 3. Moe OK, Abildskov JA: Atrial fibrillation as a self-sustaining arrhythmia independent of focal discharge. Am Heart J 5659-70, 1969 4. McWHliam JA: Fibriltar contraction of the heart. J Physiol 8: 296-310, 1887 5. Porter WT: On the results of ligation of the coronary arteries. J Physiol (London) 15:121-138, 1894 6. Gsrrey WE: The nature of fibrillatory contraction of the heartits relation to tissue mass and form. Am J Physiol 33:397-414, 1914 7. Mower MM, Mlrowski M, Spear JF, et al: Patterns of ventricular activity during catheter defibrillation. Circulation 49:858-861,

44

July 1975

The Amerkan

Journal of CARDIOLOGY

Volume 36

1974 8. Wiggers CJ: The physiologic basis for cardiac resuscitation from ventricular fibrillation-method for serial defibrillation. Am Heart J 201413-422, 1940 9. Nachlas MM, Blx HH, Mower MM, et al: Observations on defibrillators, defibrillation and synchronized countershock. Prog Cardiovasc Ois 9:64-89, 1966 10. Gulnn GA, Tacker WA, Fteyes LA, et al: Quantitation of required energy for direct human cardiac defibrillation. Circulation 50: Suppl lll:lll-229. 1974 11. Mlrowskl M, Mower MM, Staewen MS, et al: The development of the transvenous automatic defibrillator. Arch Intern Med 129: 773-779, 1972 12. Mlrowskl M, Mower MM: Personal communication, February 1974