The 1, 2, 3, 4 Phenomenon

Transcription

1 223 The 1, 2, 3, 4 Phenomenon Atrioventricular Nodal Gap in the Dog ANDRES R. TICZON, M.D., ANTHONY N. DAMATO, M.D., ANTONIO R. CARACTA, M.D., SUN H. LAU, M.D., AND GUSTAVUS A. BOBB SUMMARY This study confirms the facility with which the so-called 1, 2, 3, 4 phenomenon can be reproduced in intact dog hearts. When a series of three atrial premature beats (A n A,, A 4 ) were delivered following a constant A^A, drive, we demonstrated a narrow zone of A,-A, and A,-A, intervals in which A, conducted to the ventricles in the presence of A, bat not in its absence. We used His bundle and multiple atrial dectrograms to produce the phenomenon in nine of 16 dogs. Facilitation of conduction of A, (1) occurred above the His bundle, (2) occurred within narrow ranges of A,-A, and A,-A, intervals, (3) required penetration of A, iato the atrioventricular (AV) node, (4) required critically long A,-H, laterals, (5) always was associated with prolonged A<-H< intervals, and (6) was limited by atrial refractoriness. An apparent or pseado-1, 2, 3, 4 phenomeaon demonstrated in five of 16 dogs occurred with (1) latency between S, and A,, (2) a slaus nodal or atria) reentry beat, or (3) an atrial escape beat. la the absence of A,, A. could be made to conduct by preexdting the ventricle m advaace of V,. We foand no evidence for dual AV nodal pathways. Our results saggest that the underlying mechanism for the 1, 2, 3, 4 phenomenon it analogous to the phenomenon of the gap la AV conduction. DURING DETAILED studies of complex manifestations of concealed atrioventricular (AV) conduction in dog hearts, Moe et al. 1 described the following phenomenon. When a series of three atrial premature beats (A,, A s, A 4 ) was elicited after the atria had been driven at a constant cycle length (A,) a narrow zone of A,-A, and A^A, intervals was delineated wherein conduction of A 4 to the ventricles was dependent upon the presence of A,. It has since been referred to as the 1, 2, 3, 4 phenomenon. Moe et al. considered three possibilities to explain this phenomenon: supernormality, dual pathways in the AV node, and horizontal dissociation within the AV node. They concluded that facilitation of conduction of A 4 by A, most likely was due to the presence of dual AV pathways. 3 More than a decade has elapsed without confirmation of their classic description of this unique manifestation of concealed conduction. Recently we undertook to reproduce the phenomenon in intact dog hearts, using multiple atrial electrograms and His bundle recordings to further elucidate the mechanism responsible for the phenomenon. Our data, although confirming the facility with which the finding can be reproduced, also suggest that the underlying mechanism by which A, facilitates conduction of A 4 is not due to dual AV nodal pathways but is analogous to a gap phenomenon 1 in the AV node; that is, an increase in proximal delay in conduction of A t caused by the presence of A, permits distal segments to recover sufficiently to allow conduction of A«to the ventricles. Methods Studies were performed on 16 mongrel dogs (8-18 pounds), anesthetized with a-chloralose ( mg/kg, From the Cardiopulmonary Laboratory, U.S. Public Health Service Hospital, Staten Island, New York. Supported in part by National Heart and Lung Institute Project HL Address for reprints; Anthony N. Damato, M.D., Cardiology, U.S. Public Health Service Hospital, Staten Island, New York Received August 13, 1975; accepted for publication April 15, iv). The trachea was cannulated and ventilation with room air maintained by a Harvard respirator. After a midsternotomy, bilateral sympathectomy was performed by extirpation of both stellate ganglia and dorsal sympathetic chains. The pericardium was opened and the heart sus- TABLE I Representative Data from 16 Dogs Showing the Extent of the Zones of the /, 2, 3, 4 Phenomenon Experiment No Mean A,-A, (msec) A,-A, (msec) A,-A, (msec) (10) (5) (10) (10) (2) (2) (5) (20) (15) (9) A,-A.t (msec) (20) Pseudo A Pseudo B (25) (17) Pseudo A (45) (2) (5) Pseudo B Pseudo C (25) (30) (20) (21) Pseudo A - atrial latency; Pseudo B - atrial reentry or echo; Pseudo C - atrial escape beat; numbers in parentheses indicate range; dashes indicate that facilitation could not be demonstrated. Zone of A,-A, intervals at which conduction of A, was facilitated by A,. t Zone of A,-A, intervals at which the presence of A, facilitated conduction of A,. pended in a pericardia! cradle. Bipolar wire plunge electrodes were inserted into the regions of the sinus node, the high right atrium, Bachmann's bundle, the coronary sinus, the bundle of His, and the anterior right ventricular epicardium. The electrodes were used both for recording and

2 224 CIRCULATION RESEARCH VOL. 39, No. 2, AUGUST 1976 'A, AJ A3 ^ ATA A3A4M35 FIGURE I Representative example of the 1,2, 3,4 phenomenon elicited during high right atrial stimulation (experiment I). Tracings from lop to bottom in each panel are: external electrocardiographic lead II; electrograms from the regions of the sinus node (), Bachmann's bundle (BB), coronary sinus (CS), and bundle of His (); stimulus artifact (AS), and time lines (TL) at 10 and 100 msec. The same abbreviations are used for succeeding figures. A, represents the last atrial response to a series of 8 high right atrial drive beats at a cycle length of 400 msec (A 1-A,). S, delivered at 170 msec after S, produced an A,-A, interval of 200 msec. A, conducted to the ventricles (V,) with a prolonged A rh, interval of 200 msec as compared to the basic drive beat (A,-H, of 80 msec). This sequence and values remain the same in all panels. Panel A: at an A,-A, interval of 495 msec, A 4 conducted to the ventricle with a prolonged A,-H t interval of 160 msec. Panel B: when A t was brought closer to A, by 20 msec (A,-A, -475 msec) A, blocked in the atrioventricular (A V) node. Panel C: at identical A, -A, and A,-A, intervals as in panel B, A, was introduced at an A,-A, interval of 340 msec (A,-A, = 140 msec), which resulted in conduction of A,. Note that A, conducted to the ventricle with an A A -H t interval of 182 msec, which is 22 msec longer than the corresponding A t -H t value in panel A. Note also that the H-V intervals of all the conducted beats and the H,-H, intervals are constant at 35 and 320 msec, respectively. All measurements of electrograms were obtained from the tracings. L2-. BB. CS. A * 200 ' 310 A2 ji, j',.l [, HjH2-365 AS I B L2 iu_ A, A, 380 A,A 4510 c L2 A4 Blocked BB! U f CS. I 200 ' D5' A2 A3 A4 A4 Blocked A1 H, F "2 fc. I fjh T H,H AS j FIGURE 2 Demonstration of the narrow range ofa r A, intervals at which facilitation of conduction ofa t occurred during constant A,-A 4 intervals (experiment 15). In each panel, A t-a A X-A and A,-A, intervals are constant at 380,200, and 510 msec, respectively. A, is the 8th or last beat of the basic drive (A,-A, msec). Abo constant are the A,-H t, H,-V,,A rht, H r V* and H,-H, intervals at 90, 30,255,30, and 365 msec, respectively. Panel A: A, blocks in the atrioventricular (AY) node at an A,/A t interval of510 msec in the absence of A,. Panel B: the introduction of A, at an A,-A, interval of 355 msec fails to affect conduction of A,. Panels C and D: facilitation of conduction of A, occurs at A,-A, intervals of 350 msec and 340 msec. The longer A,-A, interval (panel C) results in the greater A,-H, interval (315 vs. 290 msec). At A,-A, intervals of <330 msec. A, again blocked in the A V node.

3 ATRIOVENTRICULAR NODAL GAP/Ticzon et al. 225 L2^_ WITHOUT A 3 A 1 A WITH A 3 * ^ BB ^ f, CS. 200 ' 336" H,H2-37O %l AS L 320 I ^ 52 L2P..^-^^.. BB.. ^Hocked f 200 ' V f,2 200 \ 150i 170 \ S2 S3 S4 H^-SOO ATA4-530 *f- AiA ' ' BO A4H4.28O H 1H 2"37O AC 1 goo \ L2. S, &2 240 "4 I i ^, ft ftftfwft I 180 I S, S2 S3 S4 H2H4-350 AS. I 200 I IV37O i Bol 270 FIGURE 3 Representative tracings showing the extent of A, influence upon A t conduction (experiment 15). In each panel, A, represents the last of a series of 8 alrial drive beats. The A,-A, and A X-A, intervals are constant at 380 msec and 200 msec, respectively. The A,-//,, //,- V\, H,-H,, A rh,, and H r V, intervals are all constant at 90, 30, 380, 260, and 30 msec, respectively. In addition, S, when delivered was constant at 350 msec, which produced an A t-a, interval of 350 msec. The tracings in each panel represent identical intervals without (left side) and with (right side) S,. Panel A: at an A,-A, interval of 520 msec, A, is blocked in the atrioventricular (A V) node. A t an identical A ra, interval as in the left side ofpanel A, the introduction of A, (which blocks in the A V node at an A,-A, interval of 350 msec) resulted in conduction of A, with a prolonged A t-h t interval of 300 msec. The H r H, interval measured 360 msec. Panel B: when the A,-A, interval was increased to 530 msec, in the absence of A, (left side) A 4 conducted to the ventricles with an A t-h t interval of 240 msec. The introduction of A, at an A,-A, interval of 350 msec (right side of panel B) resulted in conduction of A, with an A,-H t interval of 280 msec. The H r H t interval was 350 msec. Panel C: at an A,-A t of 620 msec, A, was conducted to the ventricle with A,-H, of 120 msec and the resultant H r H, interval was 280 msec (left side). Note in the right side of panel C that the A {-H, and H,-H, intervals remained unaltered when A, was applied, indicating that at this relatively long A,-A 4 interval, the presence of A, had no influence on the conduction of A,. stimulation. Selected bipolar electrograms, lead II of the electrocardiogram, and time lines generated at 10 and 100 msec were simultaneously displayed on a multichannel oscilloscope (Electronics for Medicine, model DR 12) and recorded on a magnetic tape (tape speed, 3% inches/sec). Data were subsequently transferred to photographic paper at a paper speed of 150 mm/sec. Measurements of intervals between electrograms were made with reference to the atrial electrogram in the His bundle tracings. Stimulation was accomplished by use of a programmable digital stimulator that delivered rectangular pulses 1.5 msec in duration and twice diastolic threshold in amplitude. PACING TECHNIQUES 1. The high right atrium between the sinus node and Bachmann's bundle was paced at a basic cycle length (Ai-A,). After every 8th paced beat, a premature atrial depolarization (A,) was elicited at progressively decreasing A,-A, intervals up to the point at which A, did not conduct to the ventricles. The A,-A, interval was then increased by increments of I msec to the point at which A, consistently conducted to the ventricles with the longest obtainable Aj-H, interval. The A,-A, interval was kept constant at this value.

4 226 CIRCULATION RESEARCH VOL. 39, No. 2, AUGUST 1976 o CO e 200 a? f A,A 4 msec FIGURE 4 Representative graph relating A t-h t intervals (ordinate) to different A,-A t intervals (abscissa) in the presence and absence of A t. In this experiment (no. 15) the A,-A u A t -A, and A,-A, intervals were constant at 380, 200, and 330 msec, respectively. The open circles represent conduction of A, in the absence of A, and the solid circles in the presence of A,. The solid triangles denote constant A,-H, intervals over the full range ofa t -A t intervals. At A,-A 4 intervals between 620 and 640 msec, conduction of A,, in terms of A t -H t intervals, was unaffected by A,. At A,-/4 4 intervals between 600 and 525 msec, A t -H t intervals were always longer in the presence of A, than in its absence (indicated by the vertical dashed lines between open and closed circles). The increase in A,-H, intervals in the resence of A, was as much as 50 msec. At A,-A t intervals of 520 to 490 msec (arrow) conduction of A, occurred only in the presence of A,. Note that during the zone of facilitation, the plot of the A 4 -H t intervals represents a continuum of those occurring at shorter A ra, intervals, indicating that the effects of concealed conduction by A, were linear. 2. Another premature atrial depolarization, designated A 4, was introduced at the longest A,-A 4 interval at which A 4 consistently failed to conduct to the ventricles. Under these conditions the entire A,-A 4 interval was scanned by still another premature atrial depolarization (A,), the timing of which was changed by increments or decrements of 2-5 msec. 3. Within critical zones of A,-A, (or Ai-At) intervals conduction of A«to the ventricles occurred in the presence of A, but not in its absence. 4. At constant A,-Aj intervals, the Ai-A 4 interval was progressively decreased by 2-5 msec until A, failed to conduct to the ventricle even in the presence of A,. This allowed definition of the earliest zone of facilitation of conduction of A, by A, at a given A,-A a interval. At the same constant A,-A intervals, the A,-A 4 intervals were progressively increased by 2-5 msec, and for each increment Aj was omitted and then reintroduced. The A,-A 4 interval at which A 4 conducted to the ventricles in the absence of A, defined the outer limit of facilitation by A,. The Ai-A 4 interval was continuously increased by 5- to 10-msec increments, with and without A, during each increment, until conduction of A 4 to the ventricle was unaffected by A,. 5. At a constant Ai-A 4 interval, within the zone of facilitation, the A,-A, interval was incrementally increased and decreased by 2-msec intervals to determine the effect of the timing of A, on conduction of A In three dogs the sequences of atrial stimulation were repeated at another atrial site, i.e., the coronary sinus. 7. In five dogs, during the A,-A 4 intervals at which A 4 conduction to the ventricle was facilitated by the presence of Aj, the right ventricle was preexcited before H, to determine whether, in the absence of A», conduction of A 4 to the ventricles would occur because of early retrograde depolarization of some portions of the AV node. Results The results of all 16 experiments are summarized in Table 1. In nine of 16 experiments there was facilitation of conduction of A 4 in the presence of A,, but not in its absence, at two different basic drive rates. Facilitation of conduction of A 4 did not occur in seven dogs; five of these demonstrated an apparent or pseudofacilitation of A 4. Figure 1 is a representative example of facilitation of conduction of A 4 by A,. Panel A depicts the shortest Ai-A , 280. t msec A- A msec I..1 '/i i r i i i ' A,A 3 msec FIGURE 5 Representative graph depicting the effect of different A r A, intervals (abscissa) on A,-H t interval (ordinate) during the zone of facilitation in experiment 15. At constant A X-A 4 and A,-A, intervals of 510 and 200 msec, respectively, the A,-H, interval was progressively prolonged from 260 msec to 315 msec as the A,-A, intervals were increased from 330 msec to 350 msec, indicating greater degrees of concealed conduction. The two arrows delineate the limit of the zone of A, facilitation of A x in this phase of the study.

5 ATRIOVENTRICULAR NODAL GAP/Ticzon el al. 227 i BB. AS. U 450 A. I * I V. S,S 4-4I3 A^415 Tl ^ ^^...L, I I I I I_[ j_ II I L2i. BBi AS. -^l A1A..45O 205T2 2O0T» 4 :^^- Si Si S2 S4 450 ' S, S ' 205 '110'SO' A S!S TL L11 I I.1 1 il J J..1 i i I L II I FIGURE 6 An example of apparent or pseudofacilitation due to latency between S, and its atrial response (A,) (experiment 8). In each panel, the A,-A, and A,-A, intervals are constant at 450 msec and 205 msec, respectively. Panel A: at an 5,-5 4 and A,-A 4 interval of 415 msec, A, conducted to the ventricles (V t ) in the absence of A,. Panel B: at an S,-S t and a resultant A ra, interval of 405 msec. At blocked in the atrioventricular (A V) node. Panel C: at the same 5,-5 4 interval of 405 msec as in panel B, S, was delivered at an S^S, interval of 315 msec, which resulted in an A^-A, interval of 335 msec due to a 20-msec latency between S t and its atrial response. The interposition of A, also caused an additional 20-msec latency between 5, and its atrial response (S,S, msec, A t-a t 445 msec). Thus, conduction of A, occurs because the resultant A x -A t interval was sufficiently long that it would normally conduct on its own {compare panels A and C). Latency between 5, and its atrial response (A,) persisted to the same degree in the absence of A, (not shown in the figure). interval (495 msec) at which A 4 consistently conducted to the ventricles. In panel B, the A,-A 4 interval was decreased by 20 msec to 475 msec, whereupon A 4 consistently blocked in the AV node. In panel C, the introduction of A, at an A,-A, interval of 340 msec resulted in conduction of A 4 with an A 4 -H 4 interval 22 msec greater than in panel A. The H-V intervals of all the conducted beats were constant at 35 msec. The ranges of both the A,-A, and A,-A 4 intervals during which there was facilitation of conduction of A 4 were extremely narrow, the mean values being 9 and 21 msec, respectively (Table 1). Figure 2 is an example of the narrow range of A,-A, intervals over which conduction of A 4 was facilitated. In y panels A-D, the A,-Ai, A,-A,, and A,-A 4 intervals are constant at 380, 200, and 510 msec, respectively. Only at A,-A, intervals between 350 and 340 msec did A 3 facilitate conduction of A 4 (panels C and D). At A,-A, intervals of <33O msec, A 4 again blocked in the AV node. An example of the range of A r A 4 intervals during which there was facilitation of A 4 at constant A,-A, intervals is illustrated in Figure 3. At A r A 4 intervals between 500 and 520 msec (panel A) A 4 blocked in the AV node in the absence of A,. Within this same range of Ai-A 4 intervals, A 4 conducted when A, was introduced at a constant Ai-A, interval of 350 msec (right side of panel A). The left side of panel B demonstrates that at an A r A 4 interval of 530 msec, A 4 conducted to the ventricles in the absence of A,. The right side of panel B demonstrates that, on introduction of A, at an A r A, interval of 350 msec, conduction of A 4 was maintained but occurred with a longer A 4 -H 4 interval. These findings indicate that A, penetrated the AV node and resulted in the phenomenon of concealed conduction. At A,-A 4 intervals of 620 msec and greater (panel C), the presence of A, had no effect on conduction of A 4. Figure 4 is a plot of data points from an experiment depicting the conduction characteristics of A 4 in the presence and absence of A,. At very 4ong Ai-A 4 intervals ( msec) A had no effect on conduction of A 4. This is reflected by A 4 -H 4 intervals that were essentially the same regardless of the presence or absence of Aj. At A,-A 4 intervals between 600 and 530 msec A 4 conducted with longer A 4 -H 4 intervals when A, was present. This difference was as great as 50 msec in experiments on different dogs. Facilitation of A 4 conduction (to the left of the arrows) occurred only at A,-A 4 intervals between 520 and 490 msec. Within the A,-A, zone of facilitation, conduction of A 4 (at constant A,-A 4 intervals) was, in all experiments, directly related to the A,-A, interval. An example of this relationship is illustrated in Figure 5, which shows that, as the A,-A, interval was increased from 330 to 350 msec, facilitation of conduction of A 4 occurred in association with TL -A. FIGURE 7 An example of pseudofacilitation due to a sino-atrial echo beat (experiment 13). At an 5,-5, interval of 310 msec the resultant A x -A, was 315 msec and A, blocked in the atrioventricular (A V) node. The atrium was refractory to 5 4 delivered at an Si-S, interval of 380 msec. An atrial echo beat (A,) follows A, (note high-to-low sequence of atrial activation) at an A t -A, interval of 525 msec. Conduction of A, to the ventricles simulates facilitation of conduction by A,. In the absence ofs, and A,, an Si-S, Interval of 380 msec resulted in an A i-a, interval of 390 msec at which A t blocked in the AV node (not shown).

6 228 CIRCULATION RESEARCH VOL. 39, No. 2, AUGUST 1976 longer A 4 -H 4 intervals. In the example shown, the range of increase in A t -H 4 intervals was 55 msec. In three dogs the site of atrial stimulation was changed from the high right atrium to the coronary sinus and in each case facilitation of conduction of A, by A 3 was reproducible at the two different pacing sites. In five dogs there was an apparent or pseudofacilitation of A, conduction that, on closer examination, was found to be caused by one of the following: (I) latency between S. and the A 4 response (two dogs), (2) conduction of a sinus node echo beat after block of A 4 in the AV node (two dogs), and (3) conduction of an atrial escape beat after A 3 (one dog). An example of pseudofacilitation that resulted from significant latency between S, and its A 4 response is illustrated in Figure 6. Panel A demonstrates that at S,-S 4 and A,-A t intervals of 415 msec. A, conducted normally to the ventricles in the absence of A,. In panel B, A 4 blocked in the AV node at Si-S 4 and Ai-A t intervals of 405 msec. In panel C, the S l -S < interval is the same as in panel B (405 msec) but the presence of A, caused a latency of 40 msec between S,-S 4 and their atrial responses. The resultant Ai-A 4 interval measures 445 msec and is greater than that at which A 4 normally would have conducted to the ventricles in the absence of Ai (compare panels A and C). Figure 7 illustrates an example of conduction of a sinoatrial echo beat resulting from A s at a time when the atrium was refractory to the S, stimulus. In five dogs, all of which demonstrated the 1, 2, 3, 4 phenomenon, facilitation of conduction of A, in the absence of A, could be achieved by coupling a premature ventricular beat (V,) to the preceding A. It was necessary for the A,-V. coupling interval to be such that the AV node was simultaneously depolarized antegrade and retrograde by the A, and V, impulses, respectively. This was judged to be the case when the retrograde His bundle depolarization (H,') occurred earlier than the anticipated antegrade H a (i.e., H,-H,' < H,-H,). Examples are presented in Figures 8 and 9, which show that when the AV node was depolarized both antegrade by A, and retrograde by a premature ventricular beat, conduction of A 4 was facilitated. Facilitation by a premature ventricular beat always resulted in A 4 -H 4 intervals which were not only shorter than those at which A 4 normally conducted to the ventricles on its own (Fig. 8) but also shorter than those resulting from facilitation by A, (Fig. 9). These observations are consistent with so-called "peeling back" of refractory period of the AV node. 1 Discussion The results of our study confirm previous observations made by Moe et al. 1 over a decade ago. They used only atrial and ventricular electrograms as markers and proposed that the mechanism underlying the 1, 2, 3, 4 phenomenon occurred in the AV node. The following observations from both studies support the proposal that facilitation occurs within the AV node: 1. A blocked proximal to the His bundle recording site. L2 86 AS VS A, A,-580 > jm 1M 2-4io3Si?2 M» M/A. 160,V\. fo,v 35 'H,H 2-4tt H2H AS. VS. btodwd Til I I I I I I (- A gq A H,Hj-380 ' 35 FIGURE 8 Representative example of facilitation of A, conduction by preexcitation of the ventricle (experiment II). In each panel. A, represents the 8th atrial drive beat at an A,-A i cycle length of 560 msec, and the A,-A, intervals are constant at 320 msec. Panel A: at an A,-A, interval of 705 msec A, blocked in the atrioventricular (A V) node. Note the A r H, interval of 180 msec and an H r H, interval of 410 msec. Panel B: stimulation of the ventricles (S,) at an SrS, interval of140 msec {arrow) resulted in retrograde depolarization of the bundle of His (Ht) in advance of the expected depolarization by A,. This resulted in conduction of A 4 at the same A,-A, interval as in panel A. The A,-H 4 interval is 160 msec. Note the foreshortening of the H,-H% to 380 msec as compared to an H,-H, of 410 msec in panel A. This suggest that the A V node was depolarized by both the antegrade (A t ) and retrograde (V,) impulses which resulted in a shorter time for depolarization of the entire A V node and consequently earlier recovery ("peeling back"). Panel C: when the A,-A 4 interval was increased to 720 msec, A 4 normally conducted on its own with an A 4-// 4 interval of 160 msec. Note the identical H^-H, of 410 msec as in panel A. Panel D: at an identical A t-a t interval of 720 msec as in panel C, preexcitation of ventricles (5,) at an SrS, interval of 140 msec results in a 20-msec shortening ofa t -H 4 interval (compare panels C and D). Foreshortening of A, conduction by preexcitation of the ventricles extended to A X-A, intervals of 760 msec.

7 ATRIOVENTRICULAR NODAL GAP/Ticzon et al. 229 L2 BB AS L2 B6 AS BB AS VS TL B G A^-400 A 1A 2200 A 2 I HiH 2-3S0 (rolag 1 J a/j I35I ifii lbs' H,H9-350.ft A2 H1H2-315 rnimtiiiiiiiwiuauyimnflmluiuiie^ fin A2-A4-310 A-.A..310 FIGURE 9 A comparison between the facilitating effects of A, and ventricular preexcitation on conduction of A t (experiment 9). In each panel A x represents the 8th atriaj drive at an A,-A t cycle length of 400 msec and the A,-A, interval is constant at 200 msec. Panel A: at an A,-A, interval of 510 msec A, blocked in the atrioventricular (AV) node in the absence of A,. Panel B: at the same A t-a, interval as in panel A (510 msec), the introduction of A, at an A ra, interval of 340 msec resulted in facilitation of A t conduction with an A,-H, interval of 185 msec. Panel C: at the same A ra t interval of 510 msec, preexcitation of the ventricles at an S,S, interval of 180 msec also resulted in facilitation of conduction of A t, but the A,-H t interval was 40 msec shorter than as in panel C. Note the H,-H, of 315 msec, which is shorter than the Hy-H., of 350 msec seen in the rest of the panels. This is consistent with simultaneous anlegrade and retrograde depolarization of the A V node resulting in the foreshortening of the effective refractory period of the A V node, thus allowing conduction of the subsequent A 4 (See text). 2. Block of Aj within the AV node was reflected in the fact that at A,-A 4 intervals at which A 4 normally conducted to the ventricles the phenomenon of concealed conduction (i.e., longer A 4 -H 4 intervals) was observed only in the presence of A,. 3. At short A,-A, intervals neither facilitation nor the effects of concealed conduction on A 4 were observed; this suggests that there was entrance block of A, into the AV node. Facilitation by a premature ventricular beat always resulted in A 4 -H 4 intervals that not only were shorter than those at which A 4 normally conducted to the ventricles on its own (Fig. 8) but also shorter than those resulting from facilitation by A, (Fig. 9). These observations are consistent with so-called "peeling back" of refractory period of the AV node. 1 With the exception of three critical points to be discussed below, our observations are similar to those of Moe et al. 1. The so-called 1, 2, 3,4 phenomenon occurred in 56% of our series of experiments and between an estimated 60-75% in the series by Moe et al. In only two of 16 (12%) experiments in our study was the 1, 2, 3, 4 phenomenon not demonstrated. In one case this was due to atrial fibrillation during the stimulation sequence and in the other case to an unstable preparation. 2. In the absence of A, block of A 4 occurred because the AV node had not yet recovered from the effects of A,. 3. Concealment of A, within the AV node depended upon a requisite A,-H, delay. 4. Facilitation occurred within a narrow range of A,-A, and A,-A 4 intervals (Table 1) which, in our study, were slightly shorter than those described^ Moeet ah 5. Atrial refractoriness was a limiting factor in demonstrating the phenomenon. 6. In all cases, A, not only facilitated conduction of A 4 at critical A,-A 4 intervals, but it also retarded conduction of A 4 (concealed conduction) within a range of A,-A 4 intervals during which A 4 normally conducted on its own, i.e., A 4 -H 4 intervals were always longer when A, was present and shorter when A, was omitted. Moe et al. observed this effect less frequently than we did. The effects of concealed conduction of A, were seen over a range of A,-A 4 intervals greater than 100 msec (Figs. 1, 3, 4 and 5). Outside this range of A^A, intervals, conduction of A 4 was unaffected by either the presence or absence of A, (panel D of Fig. 3). In addition, at shorter A^A, intervals, those that failed to facilitate A 4 conduction during the previously determined zone of A 4 facilitation by A,, the presence of A, produced no retarding effect on conduction of A 4 within the range of A,-A 4 intervals at which A 4 normally conducted on its own. This is consistent with the absence of concealed conduction of A, in the AV node, that is, A, either failed to enter the AV node or entered it only so slightly that it did not affect conduction of the subsequent A 4. The three differences between the two studies are the following: 1. In the present study we observed no instances of A, facilitating conduction of a preceding A,. One instance of apparent facilitation of conduction of A, by A, in reality was caused by variability in AV nodal conduction time of the basic driven beats. Thus, neither the presence nor the absence of A, (at any given A,-A, interval) had any effect on the H,-H, or V,-V, intervals. 2. Another difference was the absence of manifest AV nodal echo beats in our study. Echo beats after A, or A 4 were caused by reentry within the atrium or sinoatrial node region as evidenced by the high-to-low sequence of atrial activation (Fig. 7). The sequence of atrial activation for AV

8 230 CIRCULATION RESEARCH VOL. 39, No. 2, AUGUST 1976 nodal echo beats would be expected to be from the low-to-high atrium or similar to that seen during retrograde activation of the atrium through the AV node. 3. The third major difference was the relative frequency (five of 16) with which pseudofacilitation of A 4 was induced by one of the three mechanisms mentioned above. Moe et al. discussed three possible mechanisms to explain the 1, 2, 3, 4 phenomenon; these included supernormality, dual AV nodal conduction pathways and a phenomenon of "horizontal" dissociation within the AV node. For this latter phenomenon it was proposed that A 4, in the absence of A,, blocked in some distal portion of the AV node. In the presence of A,, which also blocked in the AV node, A 4 was delayed in the wake of A, and arrived at the distal and previously refractory portion of the AV node after it had recovered more completely. Thus, propagation to the ventricles was possible. This mechanism for facilitation of A 4 is analogous to the so-called phenomenon of gap in AV conduction.*"" Moe et al. considered the presence of dual AV nodal pathways as the most attractive hypothesis to explain the 1, 2, 3, 4 phenomenon. Their conclusion was based primarily on two observations: one was the occurrence of atrial echo beats believed to be due to AV nodal reentry and the other was the effect of Aj on AJ-VJ conduction time. They observed that, in the presence of A,, the V r V, intervals were shorter than in the absence of A,. They concluded from these findings that A s engaged a fast pathway and either arrived within a final common pathway in advance of A, or "summed" with A, within the final common pathway. In either case, the resultant V,-V 2 intervals were shorter in the presence of Aj. Conduction of A, via a fast pathway also prevented reexcitation of the AV node by an AV nodal echo beat generated from the A, response. The latter in turn resulted in a "peeling back" of AV nodal refractoriness to a degree that made propagation of A 4 possible. Against the hypothesis that a dual AV nodal pathway explains the I, 2, 3, 4 phenomenon are the following: (1) the absence of manifest AV nodal echo beats in our studies, and (2) the presence of constant A,-H s, H,-H,, and V r V, intervals throughout the entire zone of facilitation of A 4 by A,. If A, engaged a faster conduction pathway and arrived in the lower common pathway in advance of A,, the A,-H,, Hi-H,, and V r V, intervals would have been shorter in the presence of A,; in this study this never was the case. Furthermore, if Aj caused "peeling back" of AV nodal refractoriness as described above, it would be expected that, at A r A 4 intervals at which A 4 conducted on its own, the resultant A 4 -H 4 intervals would have been shorter in the presence of A,. In this study the opposite was found. However, when so-called "peeling back" of AV nodal refractoriness was produced by premature ventricular stimulation in the absence of A, (Figs. 8 and 9) the expected findings of shorter A 4 -H 4 intervals were observed. Facilitation by premature ventricular stimulation produced shorter A 4 -H 4 intervals than did facilitation by A,. Thus, it appears unlikely that "peeling back" of A V nodal refractoriness as a result of dual AV nodal pathways was the mechanism involved in facilitation of A 4 by A,. We therefore favor as the explanation for the 1, 2, 3, 4 phenomenon a mechanism referred to by Moe as horizontal dissociation within the AV node and one that has been invoked to explain certain types of gaps in AV conduction. In the absence of A,, A 4 is blocked in some part of the AV node, perhaps the distal portion. The introduction of Aj, which also blocks in the AV node, causes A 4 to be delayed in the wake of A,. During proximal delay of A 4, the distal and previously refractory portion of the AV node has sufficient time to recover. Later arrival of A 4 in the more fully recovered nodal tissue results in its propagation to the ventricles. Consistent with this proposed mechanism is the fact that the delaying effects of A, also are seen within a range of A,-A 4 intervals during which A 4 conducts on its own. Although this study has n6t proved unequivocally that the mechanism underlying the 1, 2, 3, 4 phenomenon is analogous to a gap mechanism, or the horizontal AV dissociation of Moe, the data presented speak strongly in its favor. References 1. Moe GK, Mendez C, Abildskow JA: A complex manifestation of concealed A-V conduction in tbe dog heart. Circ Res 15: 51-63, Moe GK, Preston JB, Burlington H: Physiologic evidence for dual A-V transmission system. Circ Re» 4: , Wit AL, Damato AN, Weils MB, Steiner C: Phenomenon of the gsp in atrioventricular conduction in the human heart. Circ Res 27: , Gallagher JJ, Damato AN, Caracta AR, Vargbese PJ, Josephson ME, Lau SH: Gap in A-V conduction in man; types I and II. Am Heart J 85: 78-82, Myerburg RJ, Gelband H, Hoffman BF: Functional characteristics of the gating mechanism in the canine A-V conduction system. Circ Res 28: , Akhtar M, Damato AN, Caracta AR, Batsford WP, Lau SH: The gap phenomena during retrograde conduction in man. Circulation 49: , 1974