His-Purkinje conduction during retrograde stress

Atrio-ventricular junction: Can precision electrocardiology bridge cell and electrocardiogram?

The Atrio Ventricular Junction (AVJ) is a well-defined anatomical region of the heart the physiology of which, despite extensive and numerous observations, it is not fully understood. The aim of this review is to present an up to date summary of old and more recent findings on histology, cellular electrophysiology and intracellular connectivity of this region. We have also attempted to relate our increasing understanding of nodal pathophysiology to the interpretation of the electrocardiographic (ECG) manifestations of AVN behavior. Bridging cellular observations with ECG analysis in a process we call "Precision Electrocardiology" renders this tool far more sensitive and clinically useful than the pattern analysis too often employed in the ECG interpretation.

Human His-Purkinje System: Normal Electrophysiologic Behavior

The His-Purkinje system (HPS) plays a significant role in human pathophysiology, but knowledge is scattered. This article highlights some of the relevant concepts, phenomena, and mechanisms; clarifies, expands, confirms, or modifies commonly encountered clinical events; and adds new information, which is often available but obscure. Also included are the essentials of HPS anatomy and physiology. It is important to abandon inaccurate concepts that are still taught and occasionally appear in text books.

Accessory atrioventricular node with properties of a typical accessory pathway: anatomic-electrophysiologic correlation

We report an accessory AV node producing ventricular preexcitation and comprising the retrograde limb of AV reentrant tachycardia (AVRT). A 66-year-old man presented with an anteroseptal myocardial infarction and thereafter developed recurrent, drug-refractory AVRT requiring multiple cardioversions. Electrophysiologic findings were typical for a concealed anteroseptal accessory pathway 0.5 cm anterior to the His bundle. The patient died of intractable heart failure after endocardial resection for a left ventricular aneurysm and coronary bypass grafting. Pathologic examination revealed a para-Hisian accessory AV node connecting the right atrium to ventricular myocardium immediately anterior to the His bundle at a depth of 4 mm from the endocardium. No typical AV accessory pathway was found. This is the first report of an accessory AV node that participated in AVRT. It was deeper than typical radiofrequency catheter ablation lesions.

Asymmetry of retrograde conduction and reentry within the His-Purkinje system: a comparative analysis of left and right ventricular stimulation

OBJECTIVES

The purpose of this study was to delineate retrograde His-Purkinje system conduction and reentry (V3 phenomenon) during left ventricular extrastimulation and compare them with right ventricular extrastimulation.

BACKGROUND

The V3 phenomenon has been well described in the past during right ventricular extrastimulation; however, it has not been studied systematically during left ventricular extrastimulation.

METHODS

Left and right ventricular pacing were performed in 13 patients. Retrograde and anterograde routes of impulse propagation were determined on the basis of the sequence of His (H) and right bundle (RB) potentials, H-RB intervals, as well as the QRS configuration and axis of V3 beats.

RESULTS

During right ventricular pacing, retrograde conduction of V2, when discernible, occurred exclusively through the left bundle at all coupling intervals equal to or shorter than the His-Purkinje relative refractory period, with the exception of two isolated beats. During left ventricular extrastimulation, His bundle activation was through the left bundle in nine patients and through the right or left bundle in three other patients. In one patient, the route could not be determined. The V3 phenomena occurred in eight patients during right ventricular pacing. Seven patients had a left bundle branch block pattern QRS configuration, and one had a right bundle branch block pattern configuration. V3 beats occurred in five patients during left ventricular apex pacing: left bundle branch block pattern configuration in one patient and right bundle branch block pattern configuration in four. In three of these four patients, the reentry was interfascicular and limited to the left bundle branch system.

CONCLUSIONS

The left-sided His-Purkinje system is the preferred retrograde route of impulse propagation during both left and right ventricular extrastimulation. Reentry within the His-Purkinje system elicited by right ventricular extrastimulation involves both bundle branches, whereas this reentry tends to occur within the left-sided His-Purkinje system during left ventricular pacing.

A quantitative evaluation of refractoriness within a reentrant circuit during ventricular tachycardia. Relation to termination

Programmed ventricular stimuli introduced during sustained monomorphic ventricular tachycardia frequently reset the tachycardia, resulting in a less than fully compensatory pause. A resetting response curve is generated when the set of return cycles is evaluated as the function of the coupling intervals of the extrastimuli delivered during the ventricular tachycardia. If the stimulated wave front encounters tissue within the tachycardia circuit that is not fully recovered, interval-dependent conduction changes should occur producing an increasing resetting response pattern. We quantified the magnitude of this interval-dependent conduction slowing in 17 morphologically distinct ventricular tachycardias. The slope of the increasing limb of the resetting response curve was determined by linear regression analysis and ranged from -0.30 to -1.14 (mean +/- SD, 0.70 +/- 0.25). Seven of the 17 ventricular tachycardias (41%) terminated during introduction of ventricular extrastimuli. The slope of the resetting response pattern in those ventricular tachycardias that terminated were significantly steeper than in those that did not terminate (-0.85 +/- 0.15 versus -0.61 +/- 0.21, respectively, p = 0.025). Six of the seven ventricular tachycardias terminated with programmed ventricular stimuli had a slope steeper than -0.75, whereas only one of 10 ventricular tachycardias that did not terminate exceeded this value. In conclusion, the slope of the increasing portion of the resetting response curve correlates with ability to terminate uniform sustained ventricular tachycardia by timed extrastimuli. This slope is the quantification of the magnitude of interval-dependent conduction slowing. Additionally, tissue within the reentrant circuit displaying greater degrees of interval-dependent conduction slowing may also have relatively longer effective refractory periods.

Electrophysiological determinants of antidromic reentry induced during atrial extrastimulation. Insights from a pacing model of Wolff-Parkinson-White syndrome

The electrophysiology of antidromic reentry, a less common phenomenon than orthodromic reentry, remains a poorly understood aspect of the Wolff-Parkinson-White (WPW) syndrome. We used a pacing model of ventricular preexcitation in patients without WPW, so that electrophysiological events in the normal pathway during atrial extrastimulation (A1-A2 technique) could be precisely delineated without the obscuring effect of an actual accessory pathway. Ventricular preexcitation was simulated by an A1-V1 sequential basic drive with A2-V2 extrastimulation at progressively shorter A1-A2 (equal to V1-V2) coupling intervals. At each coupling interval tested within the zone of atrioventricular (A-V) nodal effective refractory period (since anterograde block of A2 was considered mandatory for manifestation of antidromic reentry), responses were assessed after A2 alone (method I), V2 alone (method II), and A2 plus V2 (method III, the complete preexcitation model). The entire pacing protocol was performed at two A-V intervals, short (50 msec) and long (150-180 msec), thereby simulating different proximities between the A pacing site and "accessory pathway" location. Of 47 consecutive unmedicated patients screened for the study protocol, 38 failed to meet minimal prerequisites for possible initiation of antidromic reentry because of failure in 18 (38% of total) to achieve anterograde A-V nodal block of A2, even though 1:1 ventriculoatrial conduction to cycle lengths less than or equal to 500 msec (less than or equal to 400 msec in 12) was present; and poor or absent ventriculoatrial conduction in the others. The nine remaining candidates underwent the full pacing protocol. Antidromic reentry (retrograde atrial response following V2 in method III) was observed in only two cases (4% of total), and both were associated with retrograde His-Purkinje system delays (documented by method II) occurring in tandem with a long A-V interval, thereby allowing for completion of retrograde A-V nodal recovery after penetration by A2. Indeed, such a prolonged recovery time prevented initiation of antidromic reentry in six of the nine patients (proven by intact ventriculoatrial conduction in method II). Retrograde A-V nodal block of V2, independent of A2, prevented an antidromic echo in one case. Findings in our model help to clarify the various factors, including specific anterograde and retrograde A-V nodal properties; anatomic relation between the accessory and normal pathways; and the retrograde His-Purkinje system delays, that must prevail in a concerted fashion to permit the initiation of antidromic reentry during the A1-A2 technique in patients with the WPW syndrome.

Differential effects of abrupt cycle length changes on the refractoriness of accessory pathway, His-Purkinje system, atrial and ventricular myocardium in Wolff-Parkinson-White syndrome

We compared the response of the accessory pathway (AP), the atrial myocardium, the His-Purkinje system (HPS) and the ventricular myocardium during steady state (constant cycle length) and following an abrupt alteration in cycle length in 23 patients with Wolff-Parkinson-White syndrome. The durations of the anterograde and retrograde refractory periods were measured during constant drive cycle lengths of 600 and 400 ms (Method I) and during an abrupt change in cycle length of either short-to-long (400 to 600 ms) (Method II) or long-to-short (600 to 400 ms) (Method III) just before the extra stimulus. The mean durations of the anterograde effective refractory periods of the APs were 295 +/- 43, 243 +/- 39 and 273 +/- 37 ms at 600, 400 and 400 to 600 ms cycle lengths, respectively. For the atrial effective refractory periods at the three cycles, they were 238 +/- 18, 217 +/- 11 and 241 +/- 17 ms, respectively. During ventricular stimulation, the mean durations of the retrograde effective refractory periods of the APs were 263 +/- 25, 245 +/- 19 and 253 +/- 21 ms at cycle lengths of 600, 400 and 400 to 600 ms, respectively. For the relative refractory periods of the HPS, they were 335 +/- 29, 239 +/- 23 and 367 +/- 38 ms, respectively and, for the effective refractory periods of the ventricular myocardium, they were 227 +/- 17, 206 +/- 15 and 215 +/- 18 ms, respectively. The retrograde effective refractory period of the HPS could be measured in only five patients at the three cycles (600, 400 and 400 to 600 ms) and the mean values were 265 +/- 57, 225 +/- 14 and 305 +/- 27 ms, respectively. With Method III, AP and ventricular myocardium responded in a cumulative manner while HPS demonstrated paradoxical effect. Compared to Method I, changes with Methods II and III were statistically significant for all variables measured. During all three cycles, the retrograde effective refractory period of the HPS exceeded the effective refractory period of the AP; and the HPS demonstrated progressive conduction delay while the AP responded to no or minimal delays when the V1V2 intervals were similar. An abrupt cycle length change of the short-to-long type facilitated the induction of orthodromic tachycardia during ventricular pacing.(ABSTRACT TRUNCATED AT 400 WORDS)

Effect of sudden rate acceleration on the human His-Purkinje system: adaptation of refractoriness in a dampened oscillatory pattern

Although the refractoriness of the human His-Purkinje system (HPS) during constant-cycle length pacing appears to be closely related to the cycle length of the stimulation, the mode of adaptation of this refractoriness with sudden rate acceleration is not well understood. A systematic evaluation of this adaptation was performed in 14 patients with normal QRS durations and HV intervals referred for electrophysiologic evaluation. The relative refractory period of the HPS (HPS-RRP) was evaluated by the extrastimulus (S2) method during a constant ventricular drive (S1) having a cycle length as close to sinus rhythm as possible. An accelerated train of 6 ventricular beats (S1') was then added to the constant drive and the HPS-RRP of each successive beat of this train was similarly determined. Mean S1 cycle length was 750 +/- 164 msec (range 600 to 1000). Mean S1' cycle length was 475 +/- 55 msec (range 400 to 600). The HPS-RRP of each successive beat of the accelerated train was significantly shorter than that during the S1 drive and behaved in a dampened oscillatory fashion alternating from a lower value on the odd-numbered beats to a higher value on the even-numbered beats. In contrast, the effective and relative refractory periods of the ventricular myocardium during the accelerated train behaved in a cumulative manner, decreasing progressively with the first 2 beats of the train before reaching a plateau value. In conclusion, the data reported here present a new and intriguing picture of the mode of adaptation of the HPS refractoriness to sudden rate acceleration. At least in the range of the cycle lengths used in this study, the refractoriness of the HPS behaves in a dampened oscillatory manner that is radically different from the behavior of the ventricular myocardial refractoriness.

Linking: a dynamic electrophysiologic phenomenon in macroreentry circuits

The term "linking" has been used specifically to describe the mechanism for perpetuation of functional anterograde bundle branch block: namely, repetitive transseptal retrograde concealed penetration by impulses propagating along the contralateral bundle. We present selected examples that demonstrate tht linking-type phenomena actually have a wide spectrum of expression in human macroreentry circuits, particularly those incorporating either the bundle branches and His bundle or the normal pathway and Kent bundle. The examples presented are as follows: (1) persistent retrograde functional conduction delays in the His-Purkinje system during right ventricular pacing, (2) anterograde Kent bundle condution at rapid rates, dependent on prior block in the normal pathway, (3) persistent anterograde functional infra-His block of atrial impulses during rapid ventricular pacing in the presence of a retrogradely conducting accessory pathway, and (4) transient advancement of His activation with ventricular fusion complexes during overdrive ventricular pacing of bundle branch reentrant tachycardia. Based on these examples, we characterize linking as a generalized electrophysiologic phenomenon in which each successive impulse entering a macroreentry circuit propagates preferentially along one limb because of functional block in the contralateral limb resulting from the effects of the prior impulse. It is proposed that such functional block may be dynamically maintained either by repetitive impulse interference, which perpetuates local refractoriness (examples No. 1 to 3), or by repetitive impulse collision (example No. 4). The general conceptual scheme outlined can be applied to specific electrophysiologic phenomena associated with a wide variety of reentry circuits in man.

Functional characteristics of retrograde conduction in a pacing model of "endless loop tachycardia"

A pacing model was designed that stimulated "endless loop tachycardia," a complication found in the new generation of DDD (atrioventricular [AV] universal) pacemakers. The functional characteristics of the train of ventricular impulses simulating endless loop tachycardia were studied during both AV sequential pacing and basic ventricular drive. AV sequential pacing, by causing a decrease in ventriculoatrial (VA) conduction time of the first beat of the endless loop tachycardia, was associated with a decrease in the cycle length at which VA block occurred in 9 of 12 patients. The site of block was the His-Purkinje system in 4 of these 12 patients and the AV node in the remaining 8. At a cycle length with 1:1 VA conduction, a steady state VA conduction time was achieved in 2 to 4 beats (VA conduction time accommodation). The pattern of such accommodation depended on the site (His-Purkinje system versus AV node) of the maximal conduction delay. The steady state VA conduction time itself was altered with AV sequential pacing in patients showing His-Purkinje system delay, but not in patients with AV nodal delay. The results suggest that in most patients, the cycle length of VA block and the longest steady state VA conduction time will depend on the retrograde conduction time of the first beat of the tachycardia. In addition, pharmacologic measures to prevent or terminate endless loop tachycardia will have to take into account the fact that both the His-Purkinje system and the AV node can be the site of initial block.

Divergence between refractoriness of His-Purkinje system and ventricular muscle with abrupt changes in cycle length

The concept that refractoriness of the His-Purkinje system (HPS) and ventricular muscle both vary directly with cycle length is based on observations during the use of constant cycle length. During abrupt changes in ventricular cycle length, refractoriness of the ventricular muscle is known to reflect the cumulative durations of preceding cycle lengths. The effect of such changes on retrograde refractoriness of the HPS is not known. In this study refractoriness of ventricular muscle and of the HPS was evaluated in 30 patients with normal intraventricular conduction by the ventricular extrastimulus (V2) technique during constant cycle length (method I) and during abrupt cycle length changes (method II). During method II the cycle length immediately before V2 was identical to the constant cycle length of method I and therefore was designated as the reference cycle length (CLR); however, the cycle length preceding (CLP) CLR was either longer than CLR (method IIA) by 100 to 300 msec in 11 patients or shorter than CLR (method IIB) by 100 to 300 msec in 30 patients. Results showed that compared with method I, method IIA shortened the relative refractory period (RRP) of the HPS from 350 +/- 29 to 344 +/- 29 msec (p less than .04), whereas the effective refractory period (ERP) of the ventricular muscle increased from 225 +/- 21 to 233 +/- 20 msec (p less than .0001). In contrast, compared with method I, method IIB lengthened the RRP of the HPS from 335 +/- 30 to 351 +/- 35 msec (p less than .0001), whereas ERP of the ventricular muscle decreased from 223 +/- 23 to 213 +/- 22 msec (p less than .0001). Similar to the inverse relationship between CLP and RRP of the HPS, ERP of the HPS was prolonged with short CLP (method IIB) compared with long CLP (method IIA). The results indicate a marked divergence between refractoriness of the HPS and of ventricular muscle during abrupt cycle length changes; these results were not previously anticipated. Whereas ventricular muscle responded to cumulative effects of preceding cycle lengths and varied directly with CLP, the HPS appeared to respond to directional and/or dynamic changes in cycle length and varied inversely with CLP. Moreover, in contrast to ventricular muscle, the HPS appeared to be responsive to rate of change in cycle length whereby short-to-long change in cycle length had a greater effect than long-to-short change in cycle length.

Procainamide and retrograde atrioventricular nodal conduction in man

Recent studies that show a depressant effect of procainamide (PA) on retrograde conduction in patients with atrioventricular (AV) nodal reentrant tachycardia (RT) have suggested possible incorporation of AV nodal bypass tracts. Electrophysiologic effects of i.v. PA, 10 mg/kg, on retrograde AV nodal conduction were examined in 13 patients without RT, demonstrable AV nodal refractory period curves, or accessory pathways. Ventriculoatrial (VA) conduction was recorded before and after PA using intracardiac electrograms, incremental ventricular pacing and extrastimulation. With incremental pacing during the control, VA block occurred at a mean cycle length (CL) of 364.6 +/- 87.9 msec. After PA, VA conduction was abolished in five of 13 patients due to onset of retrograde block in the AV node; in seven of 13, VA block occurred at a longer paced CL after PA (344.2 +/- 51.2 msec vs 477.1 +/- 93.2 msec). In one patient, PA did not affect VA conduction. PA invariably produced prolongation in the VA interval at comparable CL of pacing. With ventricular premature stimulation, the retrograde H2A2 intervals during the control period were short (less than 50 msec) in seven of 13, intermediate (60-100 msec) in three of 13 and long (greater than 100 msec) in three of 13 cases. PA either abolished H2A2 conduction (H2 but no A2) or prolonged the H2A2 intervals by 5-20 msec in most cases in this series. The data suggest that i.v. PA almost uniformly depresses retrograde AV nodal conduction in the intact human heart. This depressant response to PA is not indicative of presence of partial or complete AV nodal bypass tracts.

Retrograde dual atrioventricular nodal pathways

Thirty-one (3.5 percent) of 887 studied patients had retrograde dual atrioventricular (A-V) nodal pathways, as manifested by discontinuous retrograde A-V nodal conduction curves (29 patients) or by two sets of ventriculoatrial (V-A) conduction intervals at the same cycle length (2 patients). All patients had A-V nodal reentrant ventricular echoes of the unusual variety induced with ventricular stimulation (25 patients had single, 2 patients had double and 4 patients had more than three ventricular echoes). The weak link of the reentrant circuit was always the retrograde slow pathway. Eleven of the 31 patients also had anterograde dual A-V nodal pathways (bidirectional dual pathways). Eight patients (26 percent) had spontaneous as well as inducible A-V nodal reentrant paroxysmal supraventricular tachycardia (of the unusual type in three and the usual type in five). In addition, three patients (10 percent) had only inducible supraventricular tachycardia (two of the unusual and one of the usual type). Retrograde dual A-V nodal pathways are uncommon. They are associated with the finding of at least single A-V nodal reentrant ventricular echoes (all patients), anterograde dual pathways (one third of patients) and A-V nodal reentrant paroxysmal supraventricular tachycardia of the usual and unusual variety (one third of patients).

Determinants of ventricular refractory periods in children with congenital heart disease: effects of cycle length and age

Ventricular effective refractory periods (ERP) and functional refractory periods (FRP) were determined by programmed ventricular extrastimulation in 53 pediatric patients with a spectrum of congenital heart disorders. Of these 63 children (ages 8 months to 18 years), 38 were preoperative, 17 had repair of their cardiac lesion via right ventriculotomy, and eight were postoperative without a ventriculotomy. We demonstrated that there was a linear relationship between the cycle length and ventricular refractory periods. The regression equations 73 + 0.29 x cycle length (msec) for the ventricular ERP (msec) and 80 + 0.30 x cycle length (msec) for ventricular FRP (msec) were found to define the determined properties of ventricular refractory periods (VRP) in children. These VRP characteristics were independent of age in children less than 13 years of age.

Role of retrograde His Purkinje block in the initiation of supraventricular tachycardia by ventricular premature stimulation in the Wolff-Parkinson-White syndrome

The precise mechanisms for paroxysmal reentrant supraventricular tachycardia (PSVT) initiation during right ventricular premature stimulation (V(2) method) were analyzed in 14 consecutive patients with Wolff-Parkinson-White Syndrome in whom the PSVT was inducible during retrograde refractory period studies. 9 patients had left-sided and the remaining 5 of 14 had right-sided ventriculo-atrial (VA) accessory pathway (AP). At the basic cycle lengths (V(1)V(1)) ranging from 550 to 900 ms (mean, 657.1+/-139.5), closely coupled V(2) (mean V(1)V(2), 357.3+/-59.2 ms, range 320-500) produced retrograde His bundle (H(2)) activation via the bundle branches and retrograde atrial (A(2)) activation via the AP. As the V(1)V(2) were further shortened, the V(2) showed a retrograde block in the His Purkinje system (HPS) and conducted to the atria via AP in 9 of 14 cases. Subsequently, the A(2) impulse conducted anterograde over the atrioventricular node-HPS to initiate a PSVT or an atrial echo response in all nine cases. In none of the patients was a PSVT induced by V(2) when the latter produced retrograde H(2) activation via the bundle branches. In 10 of 14 cases, however, the retrograde H(2) was followed by a V(3), due to macroreentry in the HPS. The V(3) in turn blocked retrogradely in the HPS while producing A(3) via the AP to initiate a PSVT or an atrial echo response in 9 of 10 cases. Retrograde block of V(2) and/or V(3) in the HPS resulted in PSVT initiation in 13 of 14 cases, whereas in the remaining 1 case the exact mechanism was not clear. In none of the patients in this series was the PSVT initiated with a retrograde block of V(2) in the atrioventricular node with or without concomitant retrograde A(2) activation via the AP. We conclude that within the ranges of cycle lengths tested, a retrograde block of V(2) and/or V(3) in the HPS is the most common mechanism for initiation of PSVT during ventricular premature stimulation in patients with the Wolff-Parkinson-White Syndrome.

Electrophysiologic study of the left ventricle: indications and safety

Eighty patients (69 with documented or suspected recurrent ventricular tachycardia or fibrillation, ten with left bundle-branch block, and one with the Wolff-Parkinson-White syndrome) underwent both right ventricular and left ventricular programmed electrical stimulation, including ventricular pacing and the introduction of one or two ventricular extrastimuli or electrode catheter mapping of the left ventricle (or both). Left ventricular catheters were introduced precutaneously via the femoral artery (of 61 patients, one required secondary repair) or via brachial arteriotomy (of 19 patients, two required secondary repair). All patients received an intravenously administered bolus of hep arin (5,000 units) following the insertion of the left ventricular catheter and then 1,000 units/hr after the first hour of study. No patients had cerebrovascular, systemic thromboembolic, or cardiac sequelae. In four (12 percent) of 34 patients with inductible ventricular tachycardia, programmed electrical stimulation of the left ventricle was required for initiation. Extensive left ventricular endocardial mapping was performed in 45 patients. Our experience suggests that (1) electrophysiologic study of the left ventricle can be performed safely, (2) programmed electrical stimulation of the left ventricle is indicated when a suspected ventricular tachyarrhythmia cannot be induced from the right ventricle, and (3) endocardial mapping of the left ventricle is indicated when surgery is being considered to abolish recurrent sustained ventricular tachycardia.

The relationship between atrioventricular nodal refractoriness and the functional refractory period in the dog

We studied the relationship between the atrioventricular nodal functional refractory period (FRP) and refractoriness by mathematical analysis and by measurement during antegrade Wenckebach cycles in 16 dogs. The FRP relates directly to the conduction time of the control beat, and inversely to the coordinates of the point on the A'-H' vs. A-A refractory curve where the slope is -1. The FRP can vary without any change in refractoriness as measured by the effective refractory period (ERP) or the refractory curve. In 16 dogs the ERP and the FRP were measured during 4:3 Wenckebach cycles. Because of changes in the control conduction times, the FRP declined and did not reflect the progressive increase in refractoriness recorded during Wenckebach cycles. The FRP is a complex parameter and does not reliably measure refractoriness.

Ventricular fibrillation during programmed ventricular stimulation: incidence and clinical implications

Ventricular fibrillation occurred in 10 (3.3 percent) of 300 patients consecutively studied with programmed ventricular stimulation. One hundred twenty-five of these patients were studied with double ventricular extrastimuli including 68 patients with and 57 patients without documented or suspected ventricular tachycardia or fibrillation, or both. Ventricular fibrillation did not develop in response to a single ventricular extrastimulus delivered during sinus rhythm, ventricular pacing or ventricular tachycardia or in response to ventricular pacing at cycle lengths of 300 msec or greater and occurred only in response to double ventricular extrastimuli. All 10 patients who manifested ventricular fibrillation during programmed stimulation were in the group of patients with suspected or documented ventricular tachycardia or fibrillation. Ventricular fibrillation was initiated in seven patients with double ventricular extrastimuli delivered during sinus rhythm or ventricular pacing and in three patients with double ventricular extrastimuli delivered during ventricular tachycardia. Four patients had spontaneous conversion to sinus rhythm and the remainder underwent defibrillation without sequelae. Recurrent ventricular fibrillation occurred clinically in 7 of the 10 patients. This study suggests that ventricular fibrillation occurs uncommonly during programmed ventricular stimulation and only in response to double ventricular extrastimuli in patients in whom spontaneous episodes are likely to occur.