Author + information
- Received April 17, 1996
- Revision received November 21, 1996
- Accepted December 20, 1996
- Published online April 1, 1997.
- ↵*Dr. Charles Fisch, Krannert Institute of Cardiology, 1111 West 10th Street, Indianapolis, Indiana 46202.
Objectives. The objective of this study was to correlate electrocardiographic (ECG) PR interval changes during normal sinus rhythm with recent observations regarding the anatomy and physiology of the dual, slow and fast atrioventricular (AV) pathways.
Background. The least common manifestation of dual AV conduction is an abrupt PR interval change in the setting of sinus rhythm. Whereas isolated cases of this phenomenon have been reported, the relatively large series we have collected makes it possible to correlate the ECG findings with the anatomy, composition and electrophysiology of the dual AV pathways.
Methods. The ECGs of 21 patients with sinus rhythm and PR interval changes consistent with dual AV node physiology were studied. Observations include duration of the short and long PR intervals, the difference between the two and the events responsible for the PR interval change.
Results. Eighteen of the 21 ECGs exhibited an abrupt and persistent PR interval change. Two of the other three ECGs manifested PR interval alternans, with slow and fast pathway, and a Wenckebach type I AV block; in the third ECG, findings compatible with simultaneous conduction along both pathways in response to a single stimulus were noted. Events responsible for the PR change included atrial premature complexes, atrial tachycardia, interpolated ventricular premature complexes and interpolated junctional premature complexes. In two the PR interval change appeared during a regular sinus rhythm.
Conclusions. The behavior of the PR interval is consistent with dual AV conduction. The PR interval duration hypothesized to represent slow pathway conduction is in keeping with the calculated anatomic length of the slow pathway. The Wenckebach type I block in the slow and fast pathways, as well as the altered conduction time in the slow pathway parallel with changing sinus rate, is evidence that the pathway is influenced by autonomic (?parasympathetic) innervation, supporting the premise that the pathways contain AV node-like tissue.
(J Am Coll Cardiol 1997;29:1015–22)
© 1997 by the American College of Cardiology
A slight difference in the rate of recovery of two divisions of the AV connexion might determine that an extra systole of the ventricle, provoked by a stimulus applied to the ventricle shortly after activity of the AV connexion, should spread up to the auricle by that part of the AV connexion having the quicker recovery process and not by the other part. In such a case, when the auricle became excited by this impulse, the other portion of the AV connexion would be ready to take up the transmission again back to the ventricle. Provided the transmission in each direction was slow, the chamber at either end would be ready to respond (its refractory phase being short) and thus the condition once established would tend to continue, unless upset by the interpolation of a premature systole. The experiments I have been able to make have given results in accord with this conclusion.
So wrote Mines () in 1913, introducing the concept of reentry and the conditions necessary for its occurrence. The experiments were conducted in the electric ray and frog by utilizing a smoked drum as the recording equipment. The postulate that the “impulse travels over different pathways on entering and reentering the depressed region” was confirmed by Schmitt and Erlanger () in 1928 and was supported by the work of other investigators demonstrating ventricular () and atrioventricular (AV) node reciprocation () in the dog heart, evidence of functional longitudinal dissociation of AV conduction ([5–9]) and differing intranodal conduction properties in response to premature atrial impulses ().
The earliest electrocardiographic (ECG) recording of reentry was published by White (), in 1915 and he () later proposed dual AV node conduction as the mechanism of the reentry. Subsequently, isolated clinical ECG tracings consistent with anterograde dual AV conduction recorded during sinus rhythm have been reported and include observations of 1) an abrupt change, either lengthening or shortening of the PR interval, and persisting for varying periods of time ([13–18]); 2) PR interval alternans ([15, 18]); 3) PR interval alternans with Wenckebach sequence of the slowly () and rapidly () conducting pathways; and 4) conduction along both pathways in response to a single sinus impulse ([20, 21]).
With the advent of intracardiac stimulation techniques, discontinuous curves with discrete “jumps” in AV conduction time, occasional reentry and reentrant tachycardia have been recorded in humans. These observations further support the concept of functional longitudinal dissociation of AV node conduction ([22–26]).
Recent observations made during catheter ablation of reentrant arrhythmias and, more specifically, identification of an anatomically discrete slow conduction pathway, thought to be located in the posteroinferior aspect of the right atrium in the area adjacent to the coronary sinus, provide strong and direct evidence for dual AV conduction ([27–30]).
Of the differing ECG manifestations of dual AV pathways, a changing PR interval during sinus rhythm, although uncommon, provides a unique model for assessing AV conduction properties. When a sufficient number of observations are available for analysis, correlative data can be derived in support of concepts relative to the anatomy and physiology of dual AV pathways. In this study we report our analyses of 21 clinical ECG tracings, gathered from our files and our published reports, that showed changing PR intervals with normal sinus rhythm, thus providing the opportunity for such correlations. To our knowledge, it is the first such study.
Of the 21 tracings showing a changing PR interval during sinus rhythm, 18 exhibited an abrupt PR interval change. In these 18 tracings, the normal PR interval was ≤200 ms in all but 1, the latter measuring 240 ms. The longer PR intervals, presumed to be reflecting conduction in the slow pathway, varied from 280 to 720 ms. In 6 tracings, the PR interval was 280 to 400 ms and in 12 it was >400 ms. The average short and long PR intervals were 179 and 465 ms, respectively, with an average difference of 286 ms. The measurements are shown in Table 1.
Events preceding an abrupt change in the PR interval included uninterrupted sinus rhythm in two patients, an atrial premature complex (APC) in two, an APC and atrial tachycardia (AT) in one, APC and ventricular premature complex (VPC) in one, a VPC in eight, an interpolated junctional premature complex (JPC) in two and electronic pacing in one. In one patient the events responsible for the PR changes were not recorded.
PR interval alternans was recorded in two patients. In one, the PR interval alternated between ∼160 and ∼240 ms with a difference of ∼80 ms. In the second instance, in addition to the alternation, there was a 3:2 Wenckebach type I block during the longer PR interval and 2:1 block during the shorter PR interval.
In one instance the findings were compatible with simultaneous conduction along two pathways with two QRS complexes in response to a single atrial impulse. The short and long PR intervals measured 280 and 560 ms, respectively, with Wenckebach type I conduction in the faster pathway.
The following figures exemplify these observations graphically and consider possible alternative mechanisms.
1.1 Spontaneous Prolongation of PR Interval (Fig. 1)
Spontaneous abrupt prolongation of the sixth PR interval from 200 to 400 ms in lead II without recognizable change of the RP interval is illustrated. In the bottom strip, the PR interval is again normal. In this instance, the abrupt prolongation of the PR interval may be due to shortening of the RP interval so small as not to be recognizable in the surface ECG. More likely, however, the change in the PR interval is due to dual pathway conduction with a shift of conduction to the slow pathway. It is possible that changes in autonomic tone altered the refractoriness or conduction, or both, of the fast pathway, shifting conduction to the slow pathway ([26, 31, 32]). Perpetuation of slow pathway conduction is more likely due to repetitive concealment of conduction from the slow pathway to the fast pathway, thus blocking conduction in the latter.
1.2 Abrupt Prolongation and Normalization of the PR Interval Due to an APC and AT (Fig. 2)
Shortening of the PR interval caused by an APC with AT is illustrated in this tracing. Sinus rhythm with a PR interval of 320 ms is interrupted by an APC and AT with shortening of the PR interval to 160 ms. The last APC recorded in the upper tracing is followed by a sinus impulse with an RP and PR interval of 560 and 340 ms, respectively. The simultaneous long RP and PR intervals prove that a short RP interval is not the cause of the long PR interval; rather, the long PR interval reflects dual AV conduction with a “jump” of conduction to the slow pathway. Repetitive concealment from the slow to the fast pathway blocks fast pathway conduction, thus perpetuating slow pathway conduction. In the bottom tracing the short PR interval follows three consecutive APCs.
1.3 Prolongation of PR Interval Induced and Terminated by an Interpolated VPC (Fig. 3)
The solid and interrupted lines identify the fast and the slow pathway conduction, respectively. The first interpolated VPC in the bottom row blocks fast pathway conduction, shifting to conduction along the slow pathway. The second interpolated VPC either blocks slow pathway conduction or prevents concealed conduction from the slow to the fast pathway, thus shifting conduction to the fast pathway.
1.4 Initiation of Conduction Along the Slow Pathway by an Interpolated VPC (Fig. 4)
This tracing is continuous. The interpolated VPC initiates conduction along the slow pathway with a PR interval longer than the PP interval; thus, the PR interval “skips” the sinus P waves. For example, in the second row all the PR intervals are longer than the PP intervals and all the sinus P waves are “skipped.” The last “skipped” P wave fails to conceal from the slow to the fast pathway, thus shifting the conduction to the fast pathway. Parallel with slowing of the sinus rate, the slow pathway conduction slows with a gradual prolongation of the PR interval from 640 ms in the second row to 840 ms in the bottom row. The measurements for this tracing, shown in Table 1(Case 17), are derived from the top strip. The delay in the slow pathway conduction concomitant with sinus slowing indicates that the slow pathway is under the influence of parasympathetic innervation and most likely contains AV node-like tissue.
1.5 PR Prolongation After an Interpolated VPC and Terminated by Reentry (Fig. 5)
Termination of slow pathway conduction by atrial reentry is illustrated in this tracing. In the top tracing a VPC initiates conduction along the slow pathway. In the bottom tracing the slow conduction is terminated by a retrograde atrial impulse inscribing an inverted P wave. This in turn is followed by ventricular reentry along the slowly conducting pathway. Failure of the latter to conceal into the fast pathway shifts conduction to the fast pathway; hence, the short PR conduction.
Manifest retrograde conduction supports concealed conduction from the slow pathway to the fast pathway as a likely mechanism of perpetuation of slow pathway conduction.
1.6 PR Prolongation Initiated by a VPC and Terminated by an APC (Fig. 6)
In the top tracing, prolongation of the PR interval follows an interpolated PVC. In the bottom tracing the PR interval normalizes after an APC. The APC conceals into the slow pathway, thus forcing conduction along the fast pathway.
The prolonged PR interval shortens gradually from ∼600 to ∼400 ms in parallel with reduction of the PP interval from ∼800 to ∼680 ms, suggesting that the slow pathway conduction is under the influence of the autonomic nervous system.
1.7 Sinus Rhythm With Alternans of PR Interval and a Paradoxic RP–PR Relation (Fig. 7)
An example of AV alternans is shown in this tracing. The paradoxic RP–PR interval relation (a long PR interval following a long RP interval) rules out shortening of the RP interval as the cause of the longer PR interval. Similarly, concealed intranodal reentry is unlikely, as indicated by the consecutive PR intervals of 160 ms in the bottom row. The latter represent fast pathway conduction time and are identical to the short PR intervals during alternans. These observations indicate that the short PR interval, with or without alternation, reflects AV conduction along the same pathway, namely, the fast pathway. Other possible, but unlikely, mechanisms responsible for the alternans include 2:1 block in the fast pathway or supernormal conduction responsible for the short PR interval.
1.8 Sinus Rhythm With PR Alternans, Type I Block in the Slow Pathway and 2:1 Block in the Fast Pathway (Fig. 8)
Sinus rhythm with PR interval alternans, paradoxic RP–PR relation, 3:2 Wenckebach type I block in the slow pathway and 2:1 block in the fast pathway are present in this tracing. The Wenckebach sequence is clearly illustrated in the top tracing beginning with the first PR interval and block of the next to last P wave. In the top row the short PR interval, reflecting conduction along the fast pathway (solid line), remains constant at ∼240 ms. The PR interval reflecting conduction along the slow pathway (interrupted line) lengthens gradually from 320 ms, the duration of the first PR interval, to 540 ms before block of the P wave. The latter would have conducted along the slow pathway. The fast pathway exhibits a 2:1 block.
1.9 Simultaneous Conduction Along the Fast and Slow Pathways (Fig. 9)
Synchronous conduction of a single atrial impulse along the fast and the slow AV pathways inscribing two QRS complexes is the least common ECG manifestation of dual AV node conduction. It has been recorded as a spontaneous event ([29, 33–35]) and less often as induced during electrophysiologic study in patients with known dual AV node conduction ([30, 36, 37]). For a single impulse to elicit two ventricular responses, conduction in the slow pathway must be sufficiently slow to allow the distant tissues to recover after excitation by an impulse conducted along the fast pathway. Similarly, there must be a unidirectional retrograde block in the slow pathway that prevents the impulse from the fast pathway from entering the slow pathway and interfering with its anterograde conduction ().
A probable example of this phenomenon is shown in this tracing. The long and short PR intervals with paradoxic RP–PR relations—namely, a short RP interval followed by a short PR interval and a long RP interval followed by a long PR interval—are best explained by dual AV conduction.
The normal PR intervals are labeled with a star. In the top row the first P wave conducts normally; the second P wave is assumed to conduct simultaneously along the fast pathway. There is a progressive prolongation of the PR interval (type I block) in the fast pathway with block of the sixth P wave in this pathway.
The third strip illustrates shift of conduction from the slow to the fast pathway (eighth P wave). Slow pathway conduction is maintained by concealed retrograde conduction from the slow to the fast pathway. The ninth P wave conducts simultaneously along the slow and the fast pathway.
The bottom strip illustrates anterograde block in the fast pathway that shifts conduction of the third P wave to the slow pathway. Failure to conceal from the slow to the fast pathway allows the fourth P wave to conduct along the fast pathway. The fifth P wave conducts along both pathways.
1.10 PR Alternans Induced by Potassium Infusion (Fig. 10)
Sinus rhythm with PR interval alternans and a paradoxic RP–PR relation induced by infusion of potassium is demonstrated. The left and right Lewis diagrams suggest concealed intranodal reentry and dual AV conduction, respectively, as two possible mechanisms for the PR alternans.
In the presence of sinus rhythm, ECG manifestations compatible with dual AV conduction (other than AV node reentrant tachycardia) are relatively uncommon and include 1) sudden and persistent prolongation or shortening of the PR interval, 2) PR alternans, and 3) dual ventricular responses to a single supraventricular impulse conducted simultaneously along the slow and fast pathways.
Alternate mechanisms for the PR changes demonstrated in the tracings presented here, although possible, are unlikely. These include supernormal conduction, prolongation of the PR interval in response to the RP shortening or junctional bigeminy, either automatic or concealed reentrant in origin. The reasoning for favoring dual pathways versus the alternative explanations is as follows:
2.1 Prolongation of PR interval due to RP interval prolongation.
To initiate conduction along the slow AV pathway, an atrial impulse must arrive during the refractory period of the fast pathway. In this study abrupt PR interval prolongation was induced by an APC, AT, interpolated JPC and VPC, ventricular pacing and, in one instance, during normal sinus rhythm without ectopic beats. Figs. 3–6illustrate slow AV pathway conduction initiated by an interpolated VPC. The VPC conducts retrogradely into the fast pathway, thus blocking anterograde conduction and forcing conduction along the slow pathway. Although it is possible that the prolonged AV conduction represents normal AV node physiology in response to the short RP interval that follows the interpolated VPC, the evidence against a short RP interval as the mechanism of slow AV conduction with prolongation of the PR interval is twofold: 1) Electrophysiologic studies in humans designed to elicit persistent prolongation of the PR interval with interpolated VPCs have met with failure. 2) If the prolonged PR interval were caused by a short RP interval, one would expect this phenomenon to be observed more often because of the relative frequency of interpolated VPCs. 3) As illustrated in Fig. 2, a prolonged RP interval is followed by a prolonged PR interval with slow pathway conduction.
To sustain conduction along a slow pathway it is necessary to invoke repetitive concealed reentry from the slow to the fast pathway and, thus, a continued refractoriness of the fast pathway. This mechanism is supported by the relatively common occurrence of manifest atrial reentry; therefore, it is reasonable to assume that concealed reentry may be responsible for continued fast pathway refractoriness and, consequently, sustained slow pathway conduction. Concealed conduction of ectopic impulses into the slow pathway will terminate the slow pathway conduction and shift conduction to the fast pathway (Fig. 3). Similarly, absence of retrograde concealment from the fast to the slow pathway is a prerequisite for slow pathway conduction.
The type I conduction delay in the slow pathway observed in Fig. 8supports the postulate that the pathway contains specialized AV node-like tissue. Similarly, changing of slow pathway conduction in parallel with that of the sinus rate suggests a vagal effect; thus the presence of specialized tissue in the slow pathway (Fig. 4and Fig. 6).
Dual AV conduction is most likely a variant of normal AV conduction with one pathway having a shorter refractory period and a longer conduction interval. That the two pathways are probably anatomic structures is suggested by termination of reentrant tachycardia with ablation of the slow or fast pathway located in the atrium. Furthermore, persistent simultaneous conduction along both pathways may result in a nonreentrant tachycardia with a manifest ventricular rate twice that of the atrial rate. Reentrant tachycardia, if present, would be relatively slow because of the participation of the slow pathway in the reentrant circuit ().
Further evidence for anatomic identity is a conduction time of the slow pathway (Table 1) estimated from the ECG that is in keeping with the calculated anatomic length and composition of the slow pathway. Assuming that the distance from anterosuperior lip of the coronary sinus orifice to the AV node is ∼1 cm (), the length of the slow pathway, and that the latter includes tissue similar to the AV node tissue, the conduction time calculated from the ECGs presented here is realistic. The conduction time of the slow pathway itself is the difference between the slow and the fast pathway conduction times. The latter reflects largely the delay in the AV node and to a lesser extent conduction along the “anatomic” fast pathway and the His bundle. The difference between slow and fast conduction time and, thus, the approximation of slow pathway conduction time is shown in Table 1.
The assumption that the slow pathway conduction is in keeping with calculated anatomic length is valid only if the slow pathway conduction is a variant of normal and similar to AV node tissue. However, should the slow pathway conduction prove a changing pathologic state, the changing velocity would result in a varying conduction time and, thus, calculation of anatomic length might be misleading.
Dual AV conduction may be recorded in the surface ECG. The most common manifestation is AV node reentrant tachycardia with anterograde slow and retrograde fast pathway conduction. Abrupt change of the PR interval, PR interval alternans and dual ventricular responses to a single supraventricular stimulus during sinus rhythm are less often observed. All suggest the presence of two functionally and anatomically distinct AV pathways. Wenckebach sequences in either the slow or the fast pathway and evidence of vagal influence suggest that the pathways comprise AV node-like tissue.
We thank Suzanne B. Knoebel, MD for review of the manuscript and Terri Kennedy for secretarial support.
↵1 Dr. Mandrola was a U.S. Public Health Service Trainee at the time of the study.
- atrial premature complex
- atrial tachycardia
- electrocardiogram, electrocardiographic
- junctional premature complex
- ventricular premature complex
- Received April 17, 1996.
- Revision received November 21, 1996.
- Accepted December 20, 1996.
- The American College of Cardiology
- Schmitt FO,
- Erlanger J
- Scherf D,
- Shookoff C
- Moe GK,
- Preston JB,
- Burlington H
- Rosenblueth A,
- Rubio R
- Mendez C,
- Han J,
- Garcia de Jalon PD,
- Moe GK
- Moe GK,
- Mendez C,
- Han J
- Mendez C,
- Moe GK
- White PD
- Katz L,
- Pick A
- ↵Pick A, Langendorf R. Interpretation of Complex Arrhythmias. Philadelphia: Lea & Febiger, 1979:227, 231.
- Surawicz B,
- Fisch C
- Fisch C
- Zipes DP
- Schuilenburg RM,
- Durrer D
- Goldreyer BN,
- Bigger JT Jr..
- Denes P,
- Wu D,
- Dhingra RC,
- Chuquimia R,
- Rosen KM
- Witt AL,
- Weiss MB,
- Berkowitz WD,
- Rosen KM,
- Steiner L,
- Damato AN
- Lee MA,
- Morady F,
- Kadish A,
- et al.
- Scheinmann MM
- Jackman WM,
- Beckman KJ,
- McClelland JH,
- et al.
- Lewis T,
- Master AM