78 CARDIAC PACING JACC Vol, 14, No.3 Septeiber189728-33 A Prospective Randomized Evaluation of Biphasic Versus Monophasic Waveform Pulses on Defibrillation Efficacy in Humans GUST H. BARDY, MD, FACC, TOM D. IVEY, MD, MARGARET D. ALLEN, MD, GEORGE JOHNSON, BSEE, RAHUL MEHRA, PxD, H. LEON GREENE, MD, FACC Seattle, Washington Biphasic waveforms have been suggested as a superior waveform for ventricular defibrillation. To test this premise, a prospective randomized intraoperative evalua- tion of defibrillation efficacy of monophasic and biphasic waveform pulses was performed in 22 survivors of out of hospital ventricular fibrillation who were undergoing car- diac surgery for implantation of an automatic dfibrilator. ‘The initial waveform used in a patient for defibrillation testing, ether monophasic or biphasic, was randomly se- lected. Subsequently, each patient served as his or her own ‘contro for defibrillation testing ofthe other waveform. The defibrillation threshold was defined as the lowest pulse amplitude that would effectively terminate ventricular fi brillation with a single discharge delivered 10 s after initiation of an episode of ventricular fibrillation induced alternating current. Esch defibrillation pulse was recorded osclloscopically, and defibrillation pulse voltage, current, resistance and stored energy were measured. Fifteen (68%) of the 22 patients had a lower defibrilla- tion threshold with the biphasic pulse, 3 (14%) had a lower threshold with the monophasle pulse and 4 (18%) had equal Aefbeilation thresholds (within 1.0 J) regardless of wave- form. The mean lading edge defibrillation threshold volt- age was 317 & 105 V when the monophasic pulse was used and 267 £ 102 V (16% less) when the biphasic pulse was used (p = 0,008). Mean leading edge defibrillation thresh- ald current was7.9 # 3.7 A when the monophasic puse was used and 6.8 + 3.8 A (14% less) when the biphasic pulse was used (p = 0.081). There were no differences in system resistance with the two waveforms: the leading edge ress- tance for the monophasic pulse was 45.3 + 17.8 and 44.1 + 144 0 for the biphasic pulse (p = 0.183). The stored nergy defbrilation threshold was 8.5 6.1 J when the ‘monophasic pulse was used and 6.3 + §.2J (26% less) when the biphasic pulse was used (p = 0.028). These findings show that in the majority of patients, biphasie pulse defibrillation is more efficient than monophasic pulse defibrillation, (J Am Coll Cardiol 1989;14:728-33) In 1947, Gurvich and Yuniev (1) demonstrated improved defibrillation efficacy of underdamped condenser discharges over standard condenser discharges in animals. In 1962, Kouwenhoven and Knickerbocker (2) proposed the use of sequential capacitor discharges of opposite polarity as a ‘means of facilitating the development of a portable defibril- lator, Subsequently, through the initial and later work of Schuder et al. (3-6) and Negovsky etal. (7), animal data had From the Division of Cardiology, Department of Medicine and the Division of Cardiothoracic Surgery, Department of Surgery from the Unive sity of Washington, Seattle, Washington. This work was supported in prt by (Grants R23-L 3617001 apd ROL-HLST4724O1A3 fom the National Insitutes of Heath, Bethesda, Maryland and by a grant ffom the American Hear Associaton, Washiegton Aflate, Seat ‘Manuscipt received October 17, 1988; revised manuscript rectved March 17, 193, accepted Apel 1,199. ‘Addcess for sepunts: Gast H. Bardy, MD, Universiy of Washington Medial Center, RO-2, 1959 Pacic Avene, Nonthes, Sete, Washingon SaI88 (01989 bythe American Collegeof Cardoogy accumulated to suggest that biphasic rectangular wave- forms, atleast of certain types, provided a higher margin of defibrillation efficacy compared with monophasic wave- forms. Preliminary work in humans by Winkle et al. (8) suggested that biphasic pulses may decrease energy require- ‘ments for defibrillation by automatic antiarrhythmic devices. If biphasic pulsing is superior to monophasic pulsing, incorporation of this methodology into an automatic defibril- lator might facilitate the development of smaller, more efficient devices. Therefore, to address the question of biphasic defibrillation efficacy in humans, we pursued a prospective randomized study in survivors of out of hospital ventricular fibrillation at the time of automatic defibrillator insertion, comparing the standard monophasic waveform pulse now used in implantable defibrillators with a biphasic waveform pulse of equal tlt. In addition, the study was designed such thatthe biphasic pulse chosen for evaluation could be generated from a single capacitor and easily imple- 73s. 10979989.50BARDY ET AL. 70 BIPHASIC VERSUS MONOPHASIC DEFIBRILLATION JACC Vol, 14, No.3 September 180728-33 ‘Table 1. Electrical Variables for Defibrlation Thresholds as a Function of Defibrillation Waveform in 22 Patients Voliage* Caren Resistance? ‘Stored Enereyé wo (amps) (ohms) CGoues) Patient No. __Momophasic Biphasic _Menophasic ‘Biphasic ——Monaphasc——=iasic —=Monophaic ‘Biphasic 1 38 T 98 o 369 382 2 Py 2 ™m as 50 18 Ma aL 23 49 3 Be co 69 6s M9 wa so 50 4 63 m 16 163 ME x09 ws ma s 1% ‘0 3A 40 ay 486 2 a 6 26 18 st 38 20 8 so as 1 om 21 86 39 as a9 99 st 8 wi ” 3 4a a8 368 4g a 9 an 22 62 a3 as 15 SL 0 w 2 32 ros sw 1 53 4 23 1% 40 “1 48 si as n n 36 61 48 0 189 a B m ns 3 389 356 2 2a “ a4 aM 94 37 BG 9 18 15 ae 2 3S a9 319 99 to 6 au 338 03 ue Mr W so 0 35 10 Iss ua ne 00 203 8 ot 39 60 a8 5.1 no 2 ” we 268 4s oy a 16 su » 31 2» m2 m3 Ba 96 1 2 1s 9 i 2. 47 2 to 2 30 70 2 so S19 19 sa Mean #SD 3721S 67S M—TYEAT BRASS ASGTINS LEM RS EOI 3282 p= 0.08; fp = 0.051; tp = 0.188; bp = 0.028 where the value of the energy (E,) stored on a capacitor is half the product ofthe capacitance (C) and the square of the voltage across the capacitor (V.) (equation 1). In this ease, the capacitor voltage is equal to the sum of the load voltage across the defibrillation electrode (V,) and the voltage across the internal resistance of the pulse generator (V)) (equation 2). If the value of the internal resistance (R,) is known, the voltage drop across the internal resistance can be calculated (equation 3). Because V,_ rather than V. was being measured, the voltage decrease (V,) across the internal resistance (R,) had to be taken into account when calculating the stored energy (equation 4). The values for V, and fy in the equation for stored energy are the leading edge values of the load voltage and current waveforms. The internal resis- tance of the pulse generator was approximately 5. and the capacitance was 120 uF. Knowledge ofthese factors allowed for calculation of stored energy values of each defibrillation pulse. Statistical considerations. Statistical analysis of the data was performed with use of paired Student's f tests to compare the efficacy of the two waveforms with respect to leading edge voltage, leading edge current, resistance and stored energy at defibrillation threshold values. The defibril- Tation threshold was defined as the lowest stored energy that could successfully terminate ventricular irillation with only ‘one discharge delivered 10s after the initiation of ventricular fibrillation. After the defibrillation threshold was determined with the intial monophasic or biphasic waveform, the de- fibrillation threshold was compared with that determined with the alternate waveform so that each patient served as his or her own control. ‘Although itis recognized that defibrillation isto a degree a statistical phenomenon, the definition of defibrillation threshold chosen for this study adjusts tothe clinical realities, and limitations of the repetitive induction of ventricular fibrillation in patients. A strength-efficacy curve would be preferred, but it would require large numbers of fibrillation defibrillation episodes to acquire, which is clinically unreal- istic. Therefore, we relied on the statistical method of the two-tailed test to accommodate forthe inherent limitations of the “defibrillation threshold.” Results Clinical and defibrillation characteristics (Table 2). Of the 22 patients studied, 16 were male and 6 were female; their mean age was 57 + 13 years. Twelve patients principally had coronary artery disease, six principally had idiopathic car- iomyopathy, one had primary electrical disease, one had long QT syndrome and two had a combination of coronary disease and cardiomyopathy. The mean ejection fraction was 0.39 + 0.18. Amiodarone had been administered over aJACC Vol 14, No.3 BARDY ET AL Bi September 199°738-33 BIPHASIC VERSUS MONOPHASIC DEFIBRILLATION Table 2. Clinical Characteristics of 22 Patents Defrilation Pest Age (yn) & Clinical Structural Heart Surpcal Hleciroge No, Genter Ahythna Disease Drugs Procedure System H 2M VF CAD. DM 0 = ‘CABG. AICD uo 2 “6F ve we 03 = ay UL 3 aM VF«G Primary clerical 089 Aniodirone ICD u sisease 4 on ve CAD, AN om = ARG, AICD uu. 5 ar VE.NSVT —ICM,MR,CAD 0 = Alc UL 5 oan ve CAD. DMI 065 = CABG, AICD cL 1 ont VE,NSVT IDC, CAD. on = Aled uo 8 Ste ve we on = leo LL ¥ aM vF lew oa = ‘led u 0 ‘tt VE.VT we 0x = ‘led LL u oM VE Iew 03 = arc LL R 0M ve CAD, OM on Amiodarone alcD B oF vex Long OT 038 - Aled syndrome or ve CAD, SEAME 099 = ARG. ale is 13M ve CAD, AML, DMI 0x ICD 16 ant ve CAD, AME 035 = CABG. patches 7 oF ve CAD, AMI 0% Alc rn ™ ved CAD. AM 025 = alco % oon ve CAD. AMI 018 CABG. AICD » ot vE CAD, AME os Amiodarone ——_AlCD: 1 xm ve Hew ase Disopyramise ICD n 70M ve CAD, DMI. LM os Amiodarone AICD Mea=S0 97213 039 = 018 ‘A= small (or average) epicardial AICD deflation path; AICD = automatic implantable cardioverter defibrillator; AMI anterior myocardial infarction: ‘CABG = coroeary patch; LY ~ left ventiolig; M = me; MR = mitral regurgitation; NSV' ‘SD = standard deviation: SEAME Tong term up to the time of defibrillation testing in four; disopyramide had been given to one of these patients. Total pulse duration for monophasic pulses was 6.2 + 1.8 s. Pulse duration was 6.1 + 1.6 ms for the intial positive phase ofthe biphasic pulse and 5.9 = 1.6 s for the negative phase of the biphasic pulse. In 19 of the 22 patients, defibrillation was performed with use of two large CPI defibrillation electrodes (one over the anterior right ventricle and one over the posterolateral left ventricle). One patient underwent defiril- lation with two small CPI defibrillation electrodes and two patients with a superior vena cava-left ventricular small patch electrode system. Defibrillation threshold data (Table 2). Fifteen (68%) of the 22 patients had a lower defibrillation threshold (> 1.0 J) with the biphasic pulse. Three patients (14%) had a lower defibrillation threshold with the monophasic pulse and four (18%) had equal defibrillation thresholds (within 1.0 J) regardless of waveform. The mean leading edge defibrillation threshold voltage was 317 + 105 V when the monophasi pulse was used and 267 + 102 V (a 16% decrease) when the biphasic pulse was used (p = 0.008) (Fig. 2) bypass surgery: CAD = coronary atery disease: DM HM » hypertrophic cardiomyopathy: ICM = idiopathic cardiomyopathy: IDC ~ idiopathic lated cardiomyopathy: L onsustaned ventricular tachycardia: S subendocardalaneiormyocarda infarction; VF = ventricular Rbilaion; iphragmatic myocardial infarction: BF = ejection faction; F = female are epicardial AICD deftriliion superie vent cava AICD coi eletrde: entcula achycadi, Mean leading edge defibrillation threshold current was 7.9 = 3.7 A when the monophasic pulse was used and 6.8 + 3.8 A (14% less) when the biphasic pulse was used (p 0.051), There were no differences in system resistances with the two waveforms. The leading edge resistance was 45.3 17.5.0 for the monophasic pulse and 44.1 + 14.4 0 for the Figure 2. Delivered leading edge voltage atthe defibrillation thresh- ld for both monophasic and biphasic pulses Yensge() o 8888 8 8732 BARDY ETAL, BIPHASIC VERSUS MONOPHASIC DEFIBRILLATION Monoonesic — Bohasic Figure3. Stored energy (voltage) fr both monophasic and biphasic pulses atthe defibrillation threshold biphasic pulse (p = 0.183). The stored energy defibrillation threshold was 8.5 + 6.1 J when the monophasic pulse was used and 6.3 + 5.2 J (a 26% decrease) when the biphasic pulse was used (p = 0.028) (Fig. 3). Discussion Clinical implications and technologie considerations. Our findings show a 26% decrease in the amount of energy needed for implantable defibrillators if biphasic pulses rather than monophasic pulses are used for defibrillation. ‘This is consistent with recent animal and human investigations 8.11.12) of biphasic pulses on defibrillation efficacy, even though the waveforms used in our study are different in shape. Although many of the recent animal studies investi- gating biphasic waveforms have used rectangular pulses rather than truncated exponential waveforms, use of rectan- gular waveforms in implantable defibrillators is impractical at this time. The capacitance necessary to deliver a rectan- gular pulse is probitive given limitations on capacitor size. Therefore, with currently available capacitors, it is more practical to use truncated pulses. We chose a 65% tilt, 120 wF, biphasic truncated expo- nential waveform to keep our investigation referrable to the monophasic pulsing techniques already in use with the automatic implantable cardioverter-defibrillator (AICD by CPI), In addition, we maintained constant waveform tlt for the monophasic pulse as well as for both phases of the biphasic pulse to control for the effect of patient to patient variability in tissue resistance on defibrillation efficacy ‘Also, the waveform tilt chosen for this study (65%) is based on previous animal data (13) showing that optimal tilt is a compromise between improved defibrillation efficacy and post-shock myocardial dysfunction at low tilts. Finally, we tried to use a biphasic pulse that could be generated by & single capacitor. The need for two capacitors can be circum- vented if the characteristics of the biphasic waveform are properly delineated. Thus, a biphasic waveform should not significantly alter device size if the pulse is created by ‘truncating the positive phase, reversing the output potarity JACE Vol, 4, No.3 September 1980-728-33 and delivering the remaining charge on the capacitor as the negative phase of the pubs. Explanations for biphasic waveform superiority. There is, no clear explanation why defibrillation with biphasic pulsing generally proves superior to defibrillation with monophasic pulsing. However, insight into the general effect of biphasic pulses on cardiac mechanical function has been provided by Jones etal. (14-17). They have shown that biphasic wave- forms with modest degrees of undershoot cause less post: shock mechanical myocardial dysfunction than monophasic pulses. Their observations are derived from a model of cultured chick embryo layered myocardial cells in which electric shocks were shown to produce transient microle- sions in the sarcolemma, possibly as a consequence of ‘membrane compression by the electric field. The size of the microlesions and the fraction of the membrane surface over which they occur increase with shock intensity. During and after the shock, an indiscriminate ionic exchange occurs through the microlesions, which results in membrane depo- larization and low resting membrane potentials (17). If, however, a small reverse polarity component is added to the biphasic pulse, Jones et al. (14-17) postulated that the phospholipids of the sarcolemma are reoriented, thereby decreasing the magnitude and duration of microlesion for- mation and, thus, membrane depolarization, Although these experiments satisfactorily account for the mechanical dys- function seen after a shock, the relation between enhanced sarcolemmal microlesion closure and enhanced defibrillation efficacy with biphasic pulses is unclear. One possible mechanism to explain why biphasic pulse defibrillation may be superior to monophasic pulse defibrit- lation is break excitation. In intracellular recordings from cells that are proximal to an extracellular stimulating elee- trode during monophasic stimulation, the cells closest to the cathode become depolarized and those underneath the an- ode become hyperpolarized (18,19). Because excitation is normally associated with a reduction in the transmembrane Potential, excitation would not be expected to occur in the cells that are hyperpolarized. However, it might be possible to explain excitation in hyperpolarized tissue by considering the phenomenon of break excitation (that is, excitation elicited at the “break” or at the end of the stimulus). This phenomenon has been observed most frequently in nerve fibers subjected to high extracellular potassium. The first explanation for this phenomenon was given by Hill (20) and subsequently by Hodgkin and Huxley (21,22), who observed that break excitation was facilitated not only in a high extracellular potassium milieu, but also in cells with low resting transmembrane potentials. In the fibrillating heart, both high extracellular potassium and low transmembrane potentials are present, thereby providing the optimal envi- ronment for break excitation to occur in the hyperpolarized tissue underneath the anode (22,23). In contrast, withJACC Vol. 16, No.3 ‘September 196072835 monophasic pulses, excitation of the hyperpolarized tissue underneath the anode may be less likely to occur. ‘An alternative explanation to break excitation for the improved defibrillation efficacy with biphasic pulsing may be associated with the two pulses contained in a “single biphasic pulse. During the positive phase of the biphasic pulse, excitation or defibrillation may occur more easily in the tissue underneath the anode (10). With the onset of the second phase of the pulse, the tissue that had been under the cathode is then under the anode. Consequently, during the second phase of the pulse, this tissue under the ‘new’ anode (that is, the old cathode) may now undergo excitation. Thus, this use of two pulses and, in effect, two anodes may facilitate depolarization of more tissue than is possible with ‘a monophasic pulse of equal leading edge voltage. ‘Conclusions. Whatever the mechanism by which biphasic pulsing facilitates defibrillation, the difference in voltage and energy requirements for defibrillation is sufficient to warrant inclusion of this waveform in future generation automat defibrillators. Together with other pulsing methods, biphasi pulsing should help decrease device size and improve de- fibrillation efficiency. References 1, Gurvieh NL, Yuniev GS. Restoration of hear rhythm during iritation bya condenser discharge. Am Rev Soviet Med 194742526, Kouventoven WB, Knickerbocker GO. The development ofa portable debate. Trans Ar lst Elec Engr 196281 (part 32831, 3, Schuder JC, Stoeckle H, Dolan AM. Transthoracic defibrillation with square wave simul one-half cycle, ore eyse, and mutcycle wave- Terms. Cite Res 19641525568 4. Schuder JC, God JH, McDaniel WC. Untrahiphbenereybyope tyre: lrom'9CR bidirectional waveform defi. Med 3:0 Eng Comput 1982:20:419-26 5. Schoder JC, Gold JH, Stoeckle H, MeDaniel WC, Cheung KN, Transtho race venticulrdefiiatin inthe 100 kg calf with syrmetcl Leycle hidiretional rectangular weave stimuli IEEE Trans Am Biomed Eng 19883041522, 6 Schuder JG, MeDaniel WC, Stoeckle H. Defibration of 100 kg calves, ‘ith asyrmesrial, bidirectional, rectangular pulses. Cardiovasc Res foes 17. Negovsky VA, Smerdov AA, Tabak VY, Venin IV, Bogushevich MS. Chien of effciency and safety of defibilating impulse. Resuscitation 05367 u 23 BIPHASIC VERSUS MONOPHASIC DEFIBRILLATION Winkle RA, Mead RH, Ruder MA, e al. Improved low enerey defi lation efficacy in man using a biphasic iruncated exponental wavelort (abst) J Am Coll Cail 1967 8:124, Lardy GH, Ivey TD, Greene HL, Evaluation of eecrode polarity on efiratin effcsy. Am J Carl I98563433-7 Barly GH, Ivey TD, Allen MD, Jobason G, Greene HL. Prospective comparison of sequenal pulse and single pulse defibation using | liferenc clinically avaiable systems in man. J Am Coll Cardia 198 M1657, ‘Warton JM, Fravier DW, Tang ASL, Ideker RE. Flecrphysiloic effects in vivo of biphasic and monophasic simul (abst). 1 Am Call, Crgit 19681114. Dixon EG, Tang ASL, Wolf PD, eal. Improved defiristion thresholds wth age contoured pian ees in biphae waveforms. Circo Tan 187.76: 17-84 Wesssle JL. Bourland 3D, Tacket WA, Geddes LA. Bipolar catheter efron in dogs using trapezokal waveforms of various ti 1 Blecroardial 1980; 1335466 Jones JL, Lepeschkin E, Jones RE, Rush S. Response of cured ryocaral cell to countershocktype elects eld stimulation. Am Physiol IESE Joes JL Jones RE. Improved defiriat waveform safety factor with biphasic waveforms, Am J Physio [983238 HO0-S Jones IL, Jones RE. Decreased defieillatr induced dysfunction with biphasic rectangular waveforms. Am J Physiol 1984247: H792-6. 2 Jones JL, Jones RE, Balasky G. Microlsion formation in myocardial calls by bighintensy elec fi stimolation. Am J Physiol 1987 sae Goto M. Brooks C McC. Membrane eacitabiliy of the fog ventricle examined by lng pss. Am I Physiol 1969;27: 1286-48 Hoshi T, Matsuda K. Excitability cycle of cardiac muscle examined by ‘racelluta stimulation, Jpn Physiol 1962124356, i AV, Excitation and secommodaton in nerve. Proe Rey Soe London 15362198 05-8, Hodgkin AL. The ionic basis of electrical activity in neve and muscle, Bl Rev I9S1:26319-409 Hodgkin AL, Huxley AF. A quantiative description of membrane curren anit application fo condition and excitation a neve] Physik I9Sts17-50-4 Janse MJ, Mosik H, Van Capelie FL, Kleber AG, Wilms Schopman F. Durer D: Ventriuar arrhythmias nthe fst 15 minutes of acute regional ‘myocarl schema in the isolated pig heart possible role of ury farens, In Kulberus HE, Welles HI. eS. Sudden Death. The Hague: Martinis Not 1889-103, Packer JO, Chong MA, West RO, Case RB. The effect of ischemia and eaton in ear ate on myocardial potassium Balance in man. Cuca [ation 197042:205-12,