The Pathophysiology of Malignant Ventricular

Arrhythmias During Acute Myocardial Ischemia

By DAVID 0. WILLIAMS, M. D., BENJAMIN J. SCHERLAG, PH.D., RONALD R. HOPE, M.D.,
NABIL EL-SHERIF, M.D., AND RALPH LAZZARA

SUMMARY
In 20 anesthetized open-chest dogs, epicardial electrograms were recorded from ischemic and non-
ischemic zones of the left ventricle during acute occlusion of the left anterior descending artery. The
average time to onset of ventricular tachycardia during atrial pacing (150-200 beats/min) was 4 min, 18 sec.
In 18 dogs, ventricular ectopic beats were induced in normal and ischemic zones after every tenth atrial
stimulus. Those induced in the ischemic zone consistently caused ventricular tachycardia earlier (mean: 3
min, 22 sec) than those in the normal zone (mean: 4 min, 11 sec) (P < 0.01). This arrhythmia, whether spon-
taneous or induced, always followed the complex which demonstrated the greatest delay of the ischemic
zone potential and increased ventricular activation time. Ventricular tachycardia was repeatedly produced
by ectopic beats with late diastolic coupling. Analysis of the episodes of tachycardia leading to fibrillation
revealed a progressive increase in the ventricular activation time of the successive beats, whereas in those
self-terminating episodes ventricular activation time progressively decreased. These data suggest that the
major determinant of malignant ventricular arrhythmias in acute ischemia may be the related abnormalities
of ventricular activation rather than the coupling of the premature ectopic beats.

Additional Indexing Wo]

Ventricular extrastimulus Heart rate Intraventricular conduction
Re-entry Left anterior descending coronary artery Refractory period

SINCE THE INTRODUCTION OF CORONARY coronary occlusion provides clear evidence of delay
Downloaded from http://ahajournals.org by on January 23, 2019

CARE UNITS the importance of the treatment and disorganization of ventricular depolarization in
and abolition of ventricular extrasystoles in prevent- the ischemic zone.' This study attempts to relate these
ing the development of life-threatening arrhythmias findings to the genesis of ectopic beats during
and thereby reducing mortality has been shown. The ischemia and to compare those spontaneous extrasys-
demonstration of a vulnerable phase during ven- toles initiating an arrhythmia with those induced by
tricular repolarization when ectopic stimulation programmed premature ventriclar stimulation. In ad-
produced tachycardia and fibrillation heightened the dition, the temporal dispersion of ventricular excita-
awareness of the malignant properties of such ectopic tion has been studied during episodes of ventricular
beats."' 2 This concept of vulnerability has been widely tachycardia which were self-terminating and com-
accepted and has proved to be of considerable clinical pared with those which progressed to fibrillation.
value. However, several reports3 4 5 have indicated
that this arrhythmogenic potential is not exclusively
confined to those extrasystoles exhibiting the R-on-T Methods
phenomenon and that tachycardia and fibrillation Twenty adult mongrel dogs were anesthetized with
may be initiated by ectopic beats occurring later in sodium pentobarbital (30 mg/kg). After intubation, ventila-
diastole. tion was maintained with room air using a Harvard
respirator. The heart was exposed through a left
The recording of surface electrograms from normal thoracotomy in the fourth intercostal space and the anterior
and ischemic areas of the canine heart during acute descending artery exposed below the origin of the anterior
septal branch.
From the Division of Cardiology, Department of Internal Two silver wires (0.012 inches diam) were inserted into
Medicine, Mount Sinai Medical Center, Miami Beach, Florida. the left vagosympathetic trunk through which stimuli (0.05
Supported in part by NIH-NHLI Contract #72-2972-M. msec duration, 20 Hz and 1-10 volts) were delivered to slow
This work was done during Dr. William's tenure of a British- the heart rate at the initiation of arrhythmia.7 Atrial pacing
American Research Fellowship of the American Heart Association with pulses of 2 msec duration, 150-200 beats/min, and
and British Heart Association. 2-10 volts was achieved by the insertion of two stainless
Address for reprints: Dr. B. J. Scherlag, Mount Sinai Medical steel wires (0.05 inches diam) into the left atrial appendage
Center, 4300 Alton Road, Miami Beach, Florida 33140. and stimulation from an S88 Grass stimulator and SIU-5
Received June 10, 1974; revision accepted for publication July 22, isolation unit.
1974. Recordings from the ventricle were made by inserting two
Circulation, Volume 50, December 1974 1163
 1164 WILLIAMS ET AL.
fine teflon-coated stainless steel wires (0.003 inches diam) formed before abnormalities persisted after release of occlu-
into the epicardium through a 25 gauge needle 11/2 inches in sion. Experiments were terminated by ventricular fibrilla-
length. The cut ends of the wires served as close bipolar tion to determine the greatest delay associated with the
recording pairs. One pair was inserted into the high lateral onset of this arrhythmia. Electrical defibrillation was not
aspect of the left ventricle to provide a control electrogram used.
and two pairs into the area supplied by the left anterior Programmed stimulation was achieved with a Medical
descending artery. Systems Devices stimulator MK. III so as to deliver an im-
In later experiments a different technique was used for pulse (2 msec duration, 5-10 volts) to the ventricle after
recording ischemic zone electrograms and was designed to every tenth atrial pacing stimuli. Stimulus intensity was
obtain information from as large an area as possible within kept constant in each dog but varied between experiments.
the ischemic zone. A large multipolar paper electrode8 was With each successive occlusion, the site of ventricular
made with two insulated silver wires (0.012 inches diam) stimulation was altered so that the effects of premature beats
which were threaded onto the surface of the paper at about induced in the nonischemic and ischemic zones could be
25-30 points to create multiple bipolar contacts with an in- compared. The atrial paced beat immediately preceding the
terpoint distance of 2-3 mm. The two wires were connected extra systole was also analyzed to indicate the progressive
to pin jack terminals and the subsequent recording effec- abnormalities that occurred when the ventricles were ac-
tively produced a "composite" electrogram from the multi- tivated along the normal atrioventricular (A-V) pathways. In
ple exposed bipolar contacts. This proved to be a convenient five dogs, the coupling interval of the ventricular stimulus,
method of recording from a large ischemic area and avoided 20% above threshold level, was adjusted in consecutive
the necessity for numerous bipolar wires. The paper elec- occlusions to fall in early and late diastole. Early diastole
trode was positioned circumferentially over the surface of was defined as 5-10 msec after the point at which the
the ischemic area and secured by fine 6-0 sutures at each stimulus was ineffective (i.e., within the T wave) and late
corner. When attached in this position, the electrode diastole as 5-10 msec before the interval at which fusion
covered areas of almost simultaneous activation, and during complexes were produced.
the control state, recorded an electrogram which was very
similar to those from bipolar wires. The only detectable Results
difference was a slight increase in duration seen in the com-
posite electrogram (see fig. 1, CIZ eg). A standard elec- A typical example of the spontaneous changes seen
trocardiographic lead and the epicardial electrograms were after occlusion of anterior descending artery is shown
continuously recorded on a Honeywell 5600 tape recorder in figure 1. Electrograms were recorded from the non-
and permanent records for analysis were subsequently ischemic zone or normal zone in the left ventricle
registered on an oscillographic-photographic recorder at
(NZeg) and from four sites within the ischemic zone
Downloaded from http://ahajournals.org by on January 23, 2019

paper speeds of 100-200 mm/sec with frequency limits of

0.1-2000 Hz and 40-200 Hz. (IZeg) supplied by this vessel. As the duration of
ischemia increased, the recorded potentials from the
Procedures subepicardial IZ sites progressively decreased in
Atrial Pacing
Recordings were taken before and during left anterior
descending artery occlusion in sinus rhythm and during CONTROL 3 MIN. 4 MIN.
atrial pacing at 150 beats/min. If significant delays in the
epicardial potentials from the ischemic zone were not ap-
parent, higher atrial pacing rates were used - up to 200 L2
beats/min. If arrhythmia occurred without apparent delay, NZeg
alternative sites within the ischemic zone were selected for
placement of the bipolar recording pairs. In each spon- IZeg i 1
taneous episode of ventricular tachycardia, the coupling in- ENDO
1 v 1-
terval of the initiating beat was determined.
ClZeg
Ventricular Premature Stimulation IZeg
Investigation with programmed ventricular stimulation IZeg
was undertaken only when the control occlusion had
produced an arrhythmia associated with appreciable epicar-
dial delay; with appropriate heart rate and vessel occlusion, NZeg 1 - -1
11..

delay of epicardial activation was achieved in all dogs. With

experience it became possible to terminate many episodes of Figure 1
ventricular tachycardia by prompt vagal-induced slowing of
the heart rate, cessation of pacing and release of occlusion, Records before (control), 3, and 4 min after coronary occlusion.
thus permitting repeated observations with a specific end Recordings from standard lead II (L2), two epicardial electrograms
point for comparative analysis. Five to ten minutes were from a nonischemic normal zone (NZ eg), two electrograms
allowed to elapse between successive occlusions and analysis recorded from bipolar wires from ischemic zones (IZ eg), one elec-
of the subsequent occlusive changes was only undertaken if trogram recorded with the composite electrode from ischemic zone
the electrograms had returned to their normal control (CIZ eg) and one endocardial ischemic zone electrogram (IZ eg
pattern recorded at the beginning of the experiment. It was ENDO). Progressive decrease in amplitude, increase in duration
found that no more than five to six occlusions could be per- and fractionation is seen in the IZ eg recordings.
Circulation, Volume 50, December 1974
 VENTRICULAR ARRHYTHMIAS IN ISCHEMIA 1165
amplitude, increased in duration, and at 4 min, had tachycardia. These results, when considered in real
fragmented into separate components. The result of time values for the period from occlusion to onset of
this delay was effectively to increase the total ven- ventricular tachycardia, were: during supraventricular
tricular activation time, as measured from the onset of rhythm (no ventricular stimulation): mean, 4 min, 18
the surface QRS to the point of maximal delay of the sec (standard deviation: ± 58 sec); normal zone, VPB:
fractionated ischemic potential. The potential mean, 4 min, 11 sec (± 66 sec); and ischemic zone
recorded from the endocardial IZ site was not affected. VPB: mean: 3 min, 22 sec (± 33 sec). Statistical
The degree of these ischemic abnormalities was un- analysis revealed no significant difference between the
equal at different sites within the ischemia area; this times for NB and NZ, VPB, but comparisons of IZ,
was found in all the dogs studied and reflects the func- VPB, with NB, and IZ, VPB, with NZ, VPB, were both
tional heterogeneity between the different recording significant at P < 0.01 level.
sites. No attempt was made to localize the sites of In three dogs, ventricular stimulation was applied,
maximum delay and limited epicardial mapping was during consecutive occlusions, to four different sites
used only to identify foci exhibiting sufficient delay for within the ischemic area. In each of these three
measurement analysis when this was not apparent animals, only one of the four sites was activated up to
from the initial electrode placements. Ventricular ac- the point at which arrhythmia resulted. Pacing at each
tivation time measured in this way is therefore almost of the other three sites produced progressively wider
certainly an underestimation since it would be un- QRS complexes until the stimulus failed to evoke a
likely that the sites of greatest delay would have been response. In all animals a site within the ischemic zone
consistently recorded. Nevertheless, it was found that could be found at which pacing was maintained to
these results were reproducible with repeated produce an arrhythmia, but the variation within this
occlusions and that the delay up to and at the onset of area probably reflected the heterogeneity of recovery
arrhythmia was remarkably consistent, provided the times.
control pattern had returned to normal after release.
It should be emphasized that the monitoring of
close bipolar electrograms, in the present study, at 80 t 1 Z VPB
multiple sites in the ischemic epicardium does not
Downloaded from http://ahajournals.org by on January 23, 2019

represent a new or radical departure from previous 701

I
studies.9' 10 Using both plunge wire bipolar or large
surface multipoint bipolar electrodes we consistently /
showed fractionation and delay of deflections 601 NZ VPB
I
recorded in the ischemic zone. Previous work from o
/ /NB
this laboratory using these techniques has c
0
501 /
/

documented delays as long as 320 msec.6 Invariably U

the greatest degree of delay coincided with the onset E
0U
4-S 40
-

/ :
of ventricular tachycardia or fibrillation. c.2
Programmed ventricular pacing was performed in Q
18 dogs. Figure 2 presents the average of the results 0.
0- 30[
for late diastolic stimulation from all the dogs studied. //:
// ,
The cumulative increase in ventricular activation time /,
20
from the preocclusion control level, represented as
zero, is plotted against time. The latter has been stan-
dardized so that the end-point represents the onset of
ventricular tachycardia; thus 1/4 T, 1/2 T and 3/4 T in-
101 / '/ _
dicate the same time point in the evolution of the L
n-
^

'AT '2AT 34T T

arrhythmia for all dogs. It can be seen that there is no Time
difference in the progressive increase in ventricular
activation time due to incremental ischemic zone Figure 2
delay between the normal beats (NB) of supraven- The incremental delay in epicardial activation (dispersion from
tricular origin and the premature beats induced in the control) is plotted against the duration of ischemia up to the onset of
normal nonischemic zone (NZ, VPB) of the ventricle. arrhythmia (T). Curves are drawn for the ventricular premature
beats induced in the ischemic zone (IZ, VPB), in the normal, non-
However, those induced by epicardial ischemic zone ischemic zone (NZ, VPB) and for the normal supraventricular beats
pacing (IZ, VPB) show greater delay with prolonga- (NB). Dispersion of ventricular activation progressively increases in
tion of ventricular activation time during the last 1/4 of each, but IZ, VPB, exhibits a greater degree of dispersion at onset of
the occlusion period up to the onset of ventricular arrhythmia.
Circulation. Volume 50, December 1974
 1166 WILLIAMS ET AL.
In all dogs studied it was found that ventricular taken 20 sec after the 3 min recording and
tachycardia followed an induced extrasystole, from demonstrates the onset of a brief period of
the normal or ischemic zone, whichever exhibited the arrhythmia, which spontaneously resolves. It can be
greatest activation delay. The total ventricular activa- seen that the CIZ electrogram of the supraventricular
tion time of the arrhythmia induced by the ex- beat becomes progressively fragmented and delays
trasystole was always greater than that of the with respect to the normal zone recording (NZ eg).
preceding extrasystole, which had evoked no Similar changes are seen in the same IZ electrogram of
response. In many cases it was apparent that as the extrasystole but are of greater magnitude. At 1
ischemia increased, an extrasystole would produce a min the delay from stimulus to the end of premature
short run of self-terminating ventricular tachycardia. ventricular activation or total ventricular activation
This pattern continued until one of the extrasystoles measures 101 msec; at 2 min, 111 msec; and at 3 min,
induced a tachycardia which persisted until abolition 136 msec. The total ventricular activation time of the
by vagal-induced atrial slowing and release of occlu- induced ectopic beat that initiates the arrhythmia is
sion, or until ventricular tachycardia progressed to 142 msec. This is not due to an increase in latency as it
fibrillation. can be seen that the interval from stimulus to onset of
Typical changes up to the onset of arrhythmia dur- the electrogram does not change as ischemia
ing ischemic zone stimulation are seen in figure 3. The progresses.
extrasystole and the atrial paced beat which im- It is interesting to note also that there is a change in
mediately precedes it are shown at 1, 2, and 3 min morphology of the extrasystole in lead 2 until it closely
after coronary artery occlusion. The lower tracing was resembles the supraventricular beat recorded im-

1 MIN 2 MIN 3 MIN

L2 _
NZeg RI -___
Downloaded from http://ahajournals.org by on January 23, 2019

-J..
CiZeg 136 :
101

SZ
STIMULUS ' 11

L
1Zeg 1V +-4 i
J, .0:' _.

L2
NZeg ~ :1
1-
-

CIZeg
142

I-
ST1MULUS

IZeg _
--- 200 l-
msec
Figure 3
The top panel shows the induced ventricular extrasystole and the preceding atrial paced beat at 1, 2, and 3 min after
occlusion. The lower panel was recorded at 3 min 20 sec. CIZ eg demonstrates progressive delay with ventricular activa-
tion time increasing from 101 msec to 136 msec after 3 min, and 142 msec at onset of arrhythmia. Note change in con-
figuration of ectopic QRS until fusion occurs. The configuration of NZ eg recording of the extrasystole changes between 2
and 3 min.
Circulation, Volume 50. December 1974
 VENTRICULAR ARRHYTHMIAS IN ISCHEMIA 1167
mediately prior to the onset of arrhythmia. The NZ eg terminating ventricular tachycardia which arose spon-
confirms that fusion has occurred. At 1 and 2 min, the taneously.
electrogram of the supraventricular beat is quite In striking contrast, the patterns of delay exhibited
different from that of the extrasystole, whereas from 3 by those episodes of ventricular tachycardia which
min on they are similar. This indicates that delay progressed to fibrillation showed progressive increase.
within the ischemic zone has progressed to such a An example is shown in figure 5. Ventricular tachycar-
degree that the normal zone is activated by the dia is induced by a premature stimulus delivered to
supraventricular pacemaker before the ventricular the normal zone and rapidly progresses to fibrillation.
stimulus is recorded. Delay in activation measured from the onset of the
This sequence was repeatedly seen and suggested a surface QRS to the most delayed of the fragmented
causal relationship. It was apparent therefore that a potentials recorded from the composite IZ elec-
ventricular extrasystole which was recorded on the trogram is 141 msec and no arrhythmia occurred. That
surface ECG late enough to produce a fusion beat was of the next extrasystole is 148 msec and arrhythmia is
still capable of initiating an arrhythmia. Since it induced. It can be seen that the small inverted poten-
appeared that the dispersion of recovery times and tial which forms the terminal component of this elec-
delay of ventricular activation of ectopic activity was trogram becomes progressively delayed until finally it
responsible for the initiation of ventricular cannot be identified with confidence. The over-all
arrhythmia, the activation time of the consecutive ventricular activation time shows a progressive in-
beats of the tachycardia were also analyzed to deter- crease from 183 msec to 271 msec. It is often difficult
mine if they were related to the outcome of the to identify the point at which tachycardia deteriorates
arrhythmia. into fibrillation and measurements of this kind rely
Figure 4 shows an example of ventricular tachycar- upon clear identification of electrographic delay
dia induced by a ventricular premature impulse which can be correlated with the appropriate surface
delivered to the ischemic zone. Recordings from an QRS complexes. Despite these limitations, all the
endocardial and epicardial site within the ischemic episodes of tachycardia progressing to fibrillation
zone are shown. Conduction delay between these two presented the similar trend of increasing delay.
sites is indicated in msec. The delay of the induced ex- The analysis of all episodes of tachycardia studied is
Downloaded from http://ahajournals.org by on January 23, 2019

trasystole which initiates the arrhythmia is 124 msec, presented in figure 6. Curve A represents the average
and that of the spontaneous beat that follows it 119 delay of those beats that led to fibrillation and B those
msec. Despite this small decrease, the arrhythmia is that spontaneously terminated prior to fibrillation. It
maintained until the delay has been considerably can be seen that in both total ventricular activation
reduced to 59 msec. This suggested that the time initially increases. In A this trend continues
maintenance of the tachycardia was dependent upon whereas in B this ultimately falls to a level below
a sustained level of delay. The first spontaneous beat, which the arrhythmia was initiated and sinus rhythm
despite the small decrease, still exhibited sufficient was regained. This finding indicates that conduction
delay to allow the arrhythmia to continue. In all ex- delay is an essential prerequisite for the initiation and
amples of this self-terminating type of arrhythmia, the maintenance of re-entry arrhythmias.
activation delay of the final beat was always less than Certainly it would appear that the changes in con-
that of the beats which had preceded it. A similar duction time during the period of study were closely
pattern was evident in those episodes of self- related to the outcome of the two types of arrhythmia.

VPZ

L2 0---O~

IZeg
F 1 vl 4,~~1
ENDO 173 97
124 119 i

CiZeg It.. I"

--A 200 I-
msec
Figure 4
Ventricular tachycardia initiated by IZ extrasystole. Conduction delay between endocardium and epicardium at IZ
recording sites is shown above arrows. After an initial increase, delay decreases and the arrhythmia terminates.
Circulation, Volume 50, December 1974
 1168 WILLIAMS ET AL.

\;L2 < _
§ 141. STIMULUS 146. 183. 214 271

i :~frIZeg

-l *- I~Zeg- ;9>>>
' ,

200
msec
Figure 5
Ventricular tachycardia is initiated by stimulus delivered to NZ. On left, the preceding extrasystole is shown, with an
activation time of 141 msec and no resulting arrhythmia. The following ten atrial paced beats are not shown. That of the
extrasystole initiating arrhythmia is 148 msec. During ventricular tachycardia, activation time progressively increases and
fibrillation occurs.

The use of a composite electrode, by effectively in- when induced by early diastolic ectopic beats (mean 3
creasing the number of bipolar contact sites within the min, 38 sec, ± 51 sec) than by late diastolic ectopic
ischemic area, enhanced the possibility of detecting beats (average 4 min, 25 sec, ± 63 sec). However, this
the most delayed potentials. Although these were not difference was of borderline significance (P = 0.075).
specifically sought, analysis of composite electrode
recordings up to and during arrhythmia yielded more Discussion
information from which the temporal sequence of ac- Wiggers1 in 1940 demonstrated the phenomenon of
Downloaded from http://ahajournals.org by on January 23, 2019

tivation delay could be deduced. Bipolar recordings ventricular vulnerability by inducing fibrillation in the
were much less consistent in demonstrating these normal dog heart with a high intensity stimulus
patterns.
It became apparent during the early part of this
study, when occlusion up to the stage of arrhythmia 240-
was observed without premature ventricular stimula- 220
tion, that the extrasystoles that initiated tachycardia - VF
200
and finally fibrillation had no specific relationship to
the so-called vulnerable period. Extrasystoles were 180
capable of producing arrhythmias irrespective of their 160
diastolic coupling time and in many examples 140
diastolic timing far removed from that part of the T 120
A
B -DS
wave which is classically considered to be vulnerable
were seen. The relationship between the R-R, Q-T, 100
and coupling intervals of ectopic beats initiating ven- 80[
tricular tachycardia is presented in table 1. In five 60-
dogs, ventricular extrasystoles induced in both normal 40
and ischemic zones at variable coupling intervals were
analyzed with respect to their arrhythmogenic proper- 20
ties. In those animals surviving repeated occlusions in 0
1~~~~~~~~~~~~
'4T Y2T 34T T
which results permitted comparative analysis, it was Time
found that arrhythmias could be initiated by ex-
trasystoles from both zones irrespective of their Figure 6
diastolic timing; indeed, ectopic ventricular activity The average delayed ventricular activation (DVA) is plotted against
arising late enough to produce fusion with the time for those episodes of ventricular tachycardia (VT) that progress
to ventricular fibrillation (VF) (curve A) and for those that ter-
supraventricular pacemakers also produced minate spontaneously (curve B). In both, delay initially increases.
arrhythmia (fig. 3). The time from occlusion to onset This continues for those progressing to VF but not for those regain-
of ventricular arrhythmias was consistently shorter ing sinus rhythm (SR).
Circulation. Volume 50, December 1974
 VENTRICULAR ARRHYTHMIAS IN ISCHEMIA 1169
Table 1 responsible for ventricular arrhythmias initiated out-
Relationship Between R-R, Q-T, and Coupling Intervals side this phase of the cardiac cycle also.
of Spontaneous and Pacing-Induced Ventricular Ectopic Marked delay of local activation should be accom-
Beats Initiating Ventricular Tachycardia panied by marked delay of recovery of excitability. It
Heart rate
was a common finding that as fragmentation and
(beats/min) R-R (msec) Q-T (msee) CI (msec) VT VF* delay occurred after coronary occlusion there was a
Spontaneoiis Ectopic Beats progressive increase in the voltage required to excite
143 420 173 268 + - local areas in the ischemic zone, even at the end of
200 300 230 295 + - diastole. Also, total failure of excitation at high levels
200 300 230 210 + - of stimulation was common. The ability to stimulate
194 310 195 280 + - adjacent sites in the ischemic zones, even with marked
194 310 195 225 + -
194 310 195 245 + - prematurity, clearly indicates a dispersion of recovery
194 310 195 285 + - of excitability at least equal to the degree of delay of
194 310 195 300 + - activation. Thus, the relative refractory period of
122 490 245 :360 + + closely adjacent tissues was markedly dispersed, a find-
158 380 190 200 + + ing which agrees with the hypothesis proposed by
179 335 210 210 + + Wiggers12 and restated by Han et al.13 that if
146 410 290 335) + +
150 400 270 315 + + "vulnerability during the relative refractory period is
1.54 390 250 340 + + due to nonuniform excitability of the tissues then
Indutced Ectopic Beats premature responses should be characterized by a
211 280 195 235 + - lower fibrillation threshold and a longer duration of
158 380 270 280 + - the vulnerable period."
157 385 290 340 + - It would appear that electrophysiological alter-
194 310 195 255) + -
ations caused by ischemia markedly increase the dura-
1 58 380 275 300 + - tion of the vulnerable phase of the ventricle. One
194 310 195 260 +
194 310 195 280 + might postulate that after coronary occlusion,
164 365 255 340 + - heterogeneous areas of delay of local activation ini-
Downloaded from http://ahajournals.org by on January 23, 2019

158 380 270 350 + + tially cause ventricular premature beats to occur and
174 345 240 240 + + these ventricular premature beats cause further dis-
150 400 260 300 + + persion of excitation initiating successive re-entrant
146 410 290 335 + +
164 365 25,5 340 + + beats or ventricular tachycardia. The ability to sustain
ventricular tachycardia, and the ultimate outcome of
*Episodes of venitricular tachyeardia terminating in veni- the arrhythmia, may be due in part to the number of
tricular fibrillation. re-entry circuits involved and the degree of delay or
Abbreviations: CI = couplinig isiterval; VT = veintricutlar block in each circuit. In a circuit showing sufficient
tachycardia; VF = ventricular fibrillatiorn.
slowing of conduction for re-entry to occur but
without block, ventricular tachycardia could sustain
itself. If another adjacent circuit became available for
delivered at or near the peak of the T wave. It has excitation from the initial circuit, further fragmenta-
since been shown that myocardial ischemia tion of the potential could occur leading to ventricular
significantly increases the degree of dispersion of fibrillation. If, however, delay and slowing of conduc-
recovery1' during this phase of the cardiac cycle. tion to the point of block occurred in a given circuit
The findings of the present study that during the prior to the connection of this circuit to another, ven-
early stages of myocardial ischemia ventricular ec- tricular tachycardia could spontaneously terminate.
topic beats falling early or late in diastole can induce An alternative explanation of these data might be
tachycardia and fibrillation appear to be at variance found in the recent reports of the effects of acute
with previous experimental and clinical evidence. It ischemia on His-Purkinje conduction and refrac-
should be noted that observations similar to ours have toriness. El-Sherif et al.'4 and Lazzara et al.15 have
been made in the clinical setting.3' 4' 5 Recently a shown that ischemia not only induces marked depres-
systematic clinical study by DeSoyza et al.5 showed sion of conduction in the His-Purkinje system but also
that the R-on-T phenomenon did not predict the oc- alters its basic responsiveness. Specifically, stimuli
currence of ventricular tachycardia in the first 24 falling well after full repolarization of ischemic cells
hours of acute myocardial infarction. Several of our either failed to produce a propagated action potential
findings indicate that a mechanism similar to that or produced poor action potentials with a long
represented by the R-on-T phenomenon may be preceding foot or prepotential. Thus, ischemia con-
Circulation, Volume 50, December 1974
 1170 WILLIAMS ET AL.

verted the "voltage"-dependent responsiveness of even when it occurs late in the cycle, is reflected in the
normal tissue into a "time"-dependent responsiveness changing QRS morphology of ventricular premature
or refractoriness. If a similar type of change occurs in beats just prior to the onset of ventricular arrhythmias.
human myocardium as a result of ischemia, stimuli This finding was always associated with increasing
falling relatively late in diastole may fail to produce a delay and fractionation of local epicardial activation
propagated response or produce a poor depolarization in the ischemic zone and is consistent with the ex-
which is conducted with marked delay. Thus the late istence of variable conduction pathways from the site
extrasystole in the ischemic myocardium may produce of excitation to the rest of the ventricle. Harris,'9 in
malignant arrhythmias due to the alteration of basic reviewing his previous work, concluded that the early
responsiveness of the tissue and not secondary to the arrhythmias are probably due to potassium liberated
conduction disturbance. This may also account for the from ischemic cells. In a recent report by Ettinger et
earlier onset of ventricular tachycardia with induced al.20 relatively local perfusion of the apical portion of
ectopic beats than during atrial pacing alone. the left ventricle was performed with isotonic KCI
through the anterior descending coronary artery.
The Electrophysiological Basis of the Malignant Within minutes of the onset of this perfusion marked
Ventricular Ectopic Beat delay and deterioration of epicardial activity was
This study indicates that in the setting of acute found concomitant with the occurrence of ventricular
ischemia in the dog the malignancy of a ventricular ectopic activity leading to ventricular tachycardia and
premature beat is not related primarily to the par- fibrillation. These ventricular arrhythmias were
ticular portion of the cardiac cycle in which it falls, characterized by coupled beating and epicardial and
but rather depends on the underlying alterations of endocardial relationships consistent with the re-entry
activation in ischemic epicardium. During the first phenomenon. It is interesting to note that Anderson et
one or two minutes of ischemia, the sinus beats and in- al.'21 working with human ventricular myocardium ex-
duced ventricular premature beats usually showed no posed to a high potassium concentration, postulated
fragmentation or delay of recorded activity in the that conduction delay and block in the myocardium
ischemic zone. However, at an average of four could give rise to QRS configurational changes.
minutes, when local ventricular activation in the
Downloaded from http://ahajournals.org by on January 23, 2019

ischemic zone began to manifest delay and fractiona- Present Concepts Concerning Genesis and Nature of Ventricular
tion, the ventricular premature beat caused greater Arrhythmias and Myocardial Ischemia
fragmentation and delay leading to ventricular Since the introduction of the concept of
tachycardia and ventricular fibrillation. Delayed ac- vulnerability by Wiggers' and the recognition of the
tivation, up to 200 msec in ischemic and infarcted R-on-T phenomenon by Smirk and Palmer2 in the
myocardium, has been previously noted by other in- clinical setting, the concept of the fibrillation
vestigators.6' 17, 18 The induced ventricular premature threshold has occupied an important place in the
beat consistently produced ventricular tachycardia literature. This concept has been extended by the
before this arrhythmia occurred spontaneously, and work of Han et al.,11 13 who proposed that the tem-
ventricular premature stimuli delivered in the poral dispersion of recovery of excitability was the
ischemic zone caused a significantly earlier predisposing condition favoring the onset of fragmen-
appearance of ventricular tachycardia and ventricular tation of activation and re-entry. However, we believe
fibrillation than those delivered in the normal zone. the concepts of fibrillation threshold and recovery of
Previous studies have shown that premature beats excitability have theoretical and practical short-
normally show dispersion of ventricular activation and comings which seriously restrict their usefulness in the
depression of recovery of excitability and that when a understanding of the origin and nature of ventricular
"premature response is invoked in an irregularly ex- arrhythmias.
citable field (ischemic zone) its propagation .. must
. First the concept of fibrillation threshold assumes
also be irregular." 13 Thus, both prematurity and the occurrence of a ventricular premature beat at an
ischemia are responsible for activation delay of the appropriately timed interval, i.e., R-on-T phenom-
premature ventricular beat, and in this respect the enon, without offering any explanation for the oc-
situation differs from that of the regularly timed con- currence of this potentially malignant extrasystole.
ducted supraventricular beat. It is not unexpected Secondly, the level of the fibrillation threshold may
therefore that this additional mechanism would result not always be a reliable indicator of the susceptibility
in greater fragmentation and delay with earlier or stability of a given heart to a lethal arrhythmia. For
appearance of an arrhythmia. example, Kent et al.22 have reported that the infusion
Another possible indication of the irregular or of nitroglycerin and phenylepheprine during myocar-
variable wavefront of the ventricular premature beat, dial ischemia increased the ventricular fibrillation
Circulation, Volume 50, December 1974
 VENTRICULAR ARRHYTHMIAS IN ISCHEMIA 1171
threshold to control, nonischemic levels. A similar than by the coupling interval of these beats.
study23 reported that 50% of the dogs receiving these At any given degree of delay, early ectopic beats ex-
drugs developed spontaneous ventricular fibrillation hibit more dispersion than those falling later in
after coronary artery ligation, whereas 92% of the diastole and therefore under these circumstances are
ischemic untreated group died with this arrhythmia. more liable to evoke an arrhythmia. However, late
Although there was a significant reduction in the in- diastolic ectopic beats will become dangerous by in-
cidence of arrhythmia in the treated group, it could be ducing fibrillation when the underlying dispersion of
inferred that the 50% who developed fibrillation ventricular activation is great. At the present time it is
would have had a normal ventricular fibrillation not possible to determine in any individual situation
threshold. This apparent inconsistency may reflect the the degree of heterogeneity of activity which has
different mechanisms responsible for ventricular developed after ischemia other than by multiple
fibrillation. Thus determination of ventricular fibrilla- direct electrode recordings. In view of this, and the
tion threshold necessitates delivery of a high intensity potential increase in dispersion which may occur as a
impulse during the "vulnerable" phase when result of further subelinical ischemia during infarc-
enhanced automaticity may be responsible for the ini- tion, it would seem prudent to ascribe potential malig-
tiation of the arrhythmia.'0 However ventricular nant properties to all ventricular ectopic beats,
fibrillation induced by spontaneous ectopic beats oc- irrespective of their diastolic timing, during the early
curring either early or late in diastole may be due to a stage of acute myocardial infarction.
re-entry mechanism.6' 8
The determination of recovery of excitability Acknowledgment
presents similar problems. Although it has been We gratefully acknowledge Mr. Jorge Rodriguez, Dr. Joseph
shown that ventricular premature beats occurring dur- Herbstman, Messrs. Israel Dingle and David Young, Jr., for their
ing myocardial ischemia cause dispersion of recovery technical assistance, Mrs. Teresa Vallone for statistical analysis, and
of ventricular excitability, the degree of such disper- Mrs. Marie Ellis for her secretarial assistance.
sion did not exceed 40 msec in the experiments
described by Han et al." This degree of nonuniform References
recovery of excitability would not be sufficient to es- 1. WIGGERS CJ, WEGRIA R: Ventricular fibrillation due to single,
Downloaded from http://ahajournals.org by on January 23, 2019

tablish re-entry since the "impulse destined to reenter localized induction and condenser shocks applied during the
the ventricle must survive for some 300 msec if it is to vulnerable phase of ventricular systole. Am J Physiol 128:
outlast the ventricular refractory period.'"24 From a 500, 1940
practical standpoint the method for determination of 2. SMIRK FH, PALMER DG: A myocardial syndrome. With
recovery of excitability requires several test particular reference to the occurrence of sudden death and
of premature systoles interrupting antecedent T waves. Am J
procedures to determine "the earliest successful S2 Cardiol 6: 620, 1960
shock artifact"" at multiple points in the ventricle. In 3. MOUNSEY P: Intensive coronary care - Arrhythmias after acute
the dynamic setting of myocardial ischemia in which myocardial infarction. Am J Cardiol 20: 475, 1967
the ischemic zone shows marked heterogeneity, such 4. STOCK JPP: In Diagnosis and Treatment of Cardiac
determinations become a formidable task even for a Arrhythmias, ed 2. London, Butterworth and Co, 1970, p
241
computer. 5. DESOYZA N, KANE J, BISSETT J, MURPHY M, DOHERTY J: Factors
The delay and fragmentation of potentials recorded predisposing to ventricular tachycardia in acute myocardial
from the ischemic area cause prolongation of the total infarction. (abstr) Clin Res 22: 4A, 1974
ventricular activation time. This increases further as 6. SCHERLAG BJ, EL-SHERIF N, HOPE R, LAZZARA R: Charac-
terization and localization of ventricular arrhythmias due to
ischemia and delay progress. The relationship myocardial ischemia and infarction. Circ Res 35: 372, 1974
between this delay of ventricular activation and the 7. VASSALLE M, GREENSPAN K, HOFFMAN BF: An analysis of
onset of ventricular arrhythmias has previously been arrhythmias induced by ouabain in intact dogs. Circ Res 13:
reported in the dog during the early phases of acute 132, 1963
myocardial ischemia.6 8 A similar mechanism, though 8. HOPE RR, WILLIAMS DO, EL-SHERIF N, LAZZARA R, SCHERLAG
BJ: The efficacy of antiarrhythmic agents during acute
not due to ischemia, has been shown to be responsible myocardial ischemia and the role of heart rate. Circulation
for the initiation of atrial fibrillation during retrograde 50: 507, 1974
conduction from the ventricle.25 Dispersion of atrial 9. HARRIS AS, ROJAs AG: The initiation of ventricular fibrillation
activation due to A-V nodal desynchronization of due to coronary occlusion. Exptl Med Surg 1: 105, 1943
retrograde conduction resulted in an increased total 10. MOE GK, HARRIS AS, WIGGERS CJ: Analysis of the initiation of
atrial activation time. The arrhythmogenic potential fibrillation by electrographic studies. Am J Physiol 134: 473,
1942
of ventricular ectopic beats during experimental acute 11. HAN J, MOE GK: Nonuniform recovery of excitability in
myocardial infarction appears to be largely deter- ventricular muscle. Circ Res 14: 44, 1964
mined by this electrophysiological mechanism rather 12. WIGGERS CJ, WEGRIA R, PINERA B: The effects of myocardial
Circulation, Volume 50, December 1974
 1172 WILLIAMS ET AL.
ischemia on the fibrillation threshold - The mechanism of 19. HARRIS AS: Potassium and experimental coronary occlusion.
spontaneous ventricular fibrillation following coronary Am Heart J 71: 797, 1966
occlusion. Am J Physiol 131: 309, 1940 20. ETTINGER PO, REGAN TJ, OLDEWURTEL HA, KHAN MI:
13. HAN J, GARCIA DE JALON PD, MOE GK: Fibrillation threshold of Ventricular conduction delay and arrhythmias during
premature ventricular responses. Circ Res 18: 18, 1966 regional hyperkalemia in the dog. Electrical and myocardial
14. EL-SHERIF N, SCHERLAG BJ, LAZZARA R, SAMET P: ion alterations. Circ Res 33: 521, 1973
Pathophysiology of tachyeardia- and bradycardia-dependent 21. ANDERSON GJ, GREENSPAN K, FIsCH C: Electrophysiologic
block in the canine proximal His-Purkinje system after acute studies on Wenckebach structures below the atrioventricular
myocardial ischemia. Am J Cardiol 33: 529, 1974 junction. Am J Cardiol 30: 232, 1972
15. LAZZARA R, EL-SHERIF N, SCHERLACG BJ: The cellular basis of 22. KENT KM, SMITH ER, REDWOOD DR, EPSTEIN SE: Beneficial
ischemic heart block and bundle branch block. Circulation electrophysiologic effects of nitroglycerin during acute
48 (suppl IV): IV-189, 1973 myocardial infarction. Am J Cardiol 33: 513, 1974
16. BASCHIERE L, PALAGI L, PULETTI M: Experimental study of the 23. BOXER JS, KENT KM, GOLDSTEIN RE, EPSTEIN SE:
pathogenesis of ventricular extrasystolia. Cardiologia 50: Nitroglycerin-induced reduction in the incidence of spon-
366, 1967 taneous ventricular fibrillation during coronary occlusion in
17. HAN J: Mechanisms of ventricular arrhythmias associated with dogs. Am J Cardiol 33: 517, 1974
myocardial infarction. Am J Cardiol 24: 800, 1969 24. CRANEFIELD PF, WIT AL, HOFFMAN BF: Genesis of cardiac
18. BOINEAL JP, Cox JL: Slow ventricular activation in acute arrhythmias. Circulation 47: 190, 1973
myocardial infarction. A source of re-entrant premature ven- 25. SCHERLAC BJ, LAZZARA R: Mechanisms of supraventricular
tricular contractions. Circulation 48: 702, 1973 tachycardia. Med Coil Va Quart 9 (1): 39, 1973
Downloaded from http://ahajournals.org by on January 23, 2019

Circulation, Volume 50, December 1974