Paradoxical Autonomic Modulation of Atrioventricular Nodal Conduction During Heart Rate Turbulence DAN WICHTERLE, IRINA SAVELIEVA, MAGGIE MEARA, A. JOHN CAMM, and MAREK MALIK From the Department of Cardiological Sciences, St. George’s Hospital Medical School, London, United Kingdom
WICHTERLE, D., ET AL.: Paradoxical Autonomic Modulation of Atrioventricular Nodal Conduction During Heart Rate Turbulence. Heart rate turbulence (HRT) represents a biphasic chronotropic response of sinus rhythm to a single ventricular premature beat (VPB). It consists of early acceleration and late deceleration of heart rate and is predominantly mediated by the autonomic nervous system. The aim of this study was to investigate if autonomic perturbations after a VPB exert a significant effect on AV conduction. Both surface ECG and the high right atrial electrogram were recorded at a sampling frequency of 1000 Hz in 26 patients (24 men, mean age 49 ± 12 years) referred for electrophysiological evaluation. The stimulation protocol consisted of series of single ventricular extrastimuli delivered from the right ventricular apex at decreasing coupling intervals. A biphasic profile of AV intervals after a single VPB was observed. The response of AV conduction to a VPB was approximately 25 times and 15 times weaker in the early and late phase, respectively, than that of R-R intervals. Thus, AV interval dynamics significantly preceded the change in R-R intervals, which is in conflict with the near to zero phase of transfer function between R-R and AV intervals described in previous studies. A significant AV turbulence was observed consisting of early shortening and later prolongation of AV intervals after VPB. Its magnitude was much smaller than that of HRT. Dynamics of AV delay has little impact on the accuracy of HRT assessment from surface ECG. The significant temporal dissociation of R-R and AV interval dynamics after a VPB remains unexplained. (PACE 2003; 26[Pt. II]:440–443) atrioventricular conduction, autonomic nervous system, heart rate turbulence, premature ventricular complex Introduction Several studies of the autonomic control of atrioventricular (AV) nodal conduction have been performed in animals1−5 and humans.6−10 Interactions among sympathetic and parasympathetic nervous activity, heart rate, and AV intervals have been studied. These studies included pacing protocols, autonomic blockade,2,3,10 blood pressure manipulation,3,6,7 cardiac reflex tests,8,9 and simple or complex analytical approaches.4,5,9,10 During normal sinus rhythm, vagal stimulation and/or sympathetic inhibition exert effects on the AV interval via two opposing mechanisms: directly by slowing AV conduction per se, and indirectly by facilitating AV conduction from slowing the heart rate. Heart rate turbulence (HRT) is a biphasic chronotropic response of sinus rhythm to a single ventricular premature beat (VPB), manifested as an early acceleration and late deceleration of heart rate,11 predominantly mediated by autonomic nerSupported in part by the Wellcome Trust Fellowship Grant No. 060683 and Research grant from Ministry of Health of Czech Republic No. 6685-3. Address for reprints: Dan Wichterle, M.D., Dept. of Cardiological Sciences, St. George’s Hospital Medical School, Cranmer Terrace, SW17 0RE London, United Kingdom. Fax: + 44-20-8725-0846; e-mail:
[email protected]
440
vous activity. Although the autonomic perturbation after a single VPB is subtle when compared with traditional provocative cardiovascular reflex tests, multiple events can be averaged to increase the accuracy of measurements. The primary objective of this study was to examine if autonomic perturbations after a VPB exert a significant effect on AV conduction (AV turbulence). The secondary objective was to establish if the changes in AV conduction might adversely affect the assessment of HRT on surface electrocardiogram (ECG). Methods The surface ECG and high right atrial (HRA) electrograms were digitally recorded at a sampling frequency of 1000 Hz in 26 patients (24 men, mean age 49 ± 12 years) referred for electrophysiological evaluations of ventricular tachyarrhythmias. The patients underwent a stimulation protocol consisting of three series of single ventricular extrastimuli delivered from the right ventricular apex every 20 seconds at decremental coupling intervals from 750 to 400 ms in 50-ms step. This protocol was completed prior to standard programmed electrical stimulation. QRS complexes and HRA potentials were automatically detected using a combination of threshold and derivative methods, manually
January 2003, Part II
PACE, Vol. 26
AUTONOMIC MODULATION OF AV CONDUCTION
inspected end edited with respect to morphological and rhythm irregularities. Their exact positions were corrected by a criterion of maximum correlation with a corresponding signal-averaged template. Dynamics of R-R and AV intervals was separately analyzed irrespective of the coupling interval for each stimulation episode, which was associated with full compensatory pause after a VPB, and was free of nonsinus beats (15 beats before and 10 beats after the captured beat). Averaged post-VPB profiles of R-R and AV intervals were constructed for individual patients. Only patients with a typical post-VPB profile of R-R intervals were selected for further analysis. Turbulence Onset (TOR-R ) was assessed as a percentage difference between means of two R-R intervals immediately before and after a VPB. Turbulence Slope (TSR-R ) was defined as the maximum positive slope of regression line assessed over a sequence of five subsequent R-R intervals starting at 3rd , 4th , or 5th beat after VPB. The same analytical approach was used to describe the dynamics of AV intervals after VPB and to obtain descriptors TOAV and TSAV . Descriptors of HRT and AV turbulence were calculated as means of all episodes in the pool of selected patients. Similarly, averaged post-VPB profiles of R-R and AV intervals were constructed for this group. Results Typical post-VPB profiles of R-R intervals were found in 7 of 7 patients with preserved left ventricular function (5 idiopathic ventricular tachyarrhythmias, 1 coronary artery disease, 1 hypertrophic cardiomyopathy), and in 7 of 19 patients with left ventricular dysfunction (ejection fraction 0.32 ± 0.05, 5 coronary artery disease, 1 dilated cardiomyopathy, 1 sarcoidosis). A total of 206 analyzable stimulation episodes were detected in these 14 patients. Immediately before single ventricular extrastimuli, R-R and AV interval were 831 ± 123 and 189 ± 32 ms, respectively. VPBs had prematurity of 84 ± 16% (means ± SD). Descriptors of HRT and AV turbulence were as follows: TOR−R − 3.2 ± 4.2%; TSR-R 14.7 ± 16.5 ms/R-R; TOAV − 0.60 ± 3.4%; TSAV 0.93 ± 1.3 ms/R-R (means ± SD). The medians, quartile range, and nonoutlier minimum/maximum values are shown in Figure 1. Averaged post-VPB profiles of R-R and AV intervals in relative numbers are shown in Figure 2, which demonstrates the biphasic response of AV intervals after a single VPB. Discussion It has been shown that early acceleration and late deceleration of heart rate after a single VPB PACE, Vol. 26
Figure 1. Box-whisker plots of descriptors of HRT and AV turbulence. Points, boxes, and whiskers represent medians, quartile ranges, and nonoutliers minimum/ maximum values, respectively. AV = atrioventricular; HRT = heart rate turbulence; TOR-R = conventional turbulence onset; TSR-R = conventional turbulence slope; TOAV = AV interval derived turbulence onset; TSAV = AV interval derived turbulence slope.
correspond to changes in systolic blood pressure (SBP) in a way fully compatible with baroreflex mechanisms.12,13 In addition, HRT can be nearly eliminated by vagal blockade with atropin.14 However, the primary cause of blood pressure dynamics after a VPB has not been fully elucidated. VPBs are associated with hemodynamically less efficient ventricular contraction, causing an initial fall in blood pressure and secondary withdrawal of
Figure 2. Averaged post-VPB profiles of R-R intervals (closed circles, bold line) and AV intervals (open diamonds, thin line) expressed as percentages of preVPB values are plotted as means ± SEM (grey and open bars). Intervals numbered −2 and −1 indicate those preceding the VPB; R-R intervals numbered 0 and 1, and AV interval numbered 0 are not quantifiable due to the coincidence with VPB. AV = atrioventricular; VPB = ventricular premature beat.
January 2003, Part II
441
WICHTERLE, ET AL.
Figure 3. Averaged post-VPB profiles of R-R intervals (closed circles, bold line) and systolic blood pressure (open circles, thin line) expressed as percentages of pre-VPB values are plotted as means ± SEM (grey and open bars). R-R intervals numbered −1 and −2 indicate those preceding the VPB; R-R intervals numbered 0 and 1 are those not quantifiable due to the coincidence with VPB. Plotted values of SBP correspond to R-R intervals in which they occurred. SBP plot is shifted slightly to the left for better legibility. Unpublished data from different study in 17 subjects with 323 episodes of spontaneous isolated VPB. SBP = systolic blood pressure; VPB = ventricular premature beat.
vagal efferent activity. The cause of late dynamics of blood pressure, characterized, at least in individuals with normal left ventricular function, by an overshoot to “positive” values (i.e., by transient peak in blood pressure several cycles after a VPB) is not clear (Fig. 3). The VPB induces a prominent burst of muscle sympathetic nerve traffic, which is observed approximately at the second post-VPB sinus beat.15 However, it is unlikely that this burst of sympathetic activity is the cause of the transient hypertension for two reasons. First, impulse response analyses of transfer function between sympathetic stimulation and heart rate in anesthetized cats has shown that the response to sympathetic stimulation is slow (50% response was attained after approximately 15 s).5 Second, sympathetic excitation after a VPB is followed by a >10-second period of sympathetic inactivity in healthy subjects, which opposes the initial sympathetic response. Therefore, the sympathetic response to a VPB is secondary to blood pressure dynamics and not vice-versa. Anecdotal evidence also indicates that heart rate and blood pressure dynamics after a VPB are not influenced by β-adrenergic blockade, which supports the hypothesis that the sympathetic arm of the autonomic nervous system is not significantly involved in the mechanism of HRT.16 It may also be hypothesized that late overshoot of blood pressure 442
after the VPB is based on a nonautonomic mechanism, representing the response to the instant decrease in cardiac output in the wake of the VPB, leading to an increase in preload and to a compensatory increase in cardiac output during the late phase. Therefore, vagal activity may be solely involved in HRT. Even using simple descriptors of the AV interval dynamics, the response of AV conduction to a VPB was approximately 25-fold weaker in the early phase and 15-fold weaker in the late phase than that of R-R intervals. This is similar to findings by others.10 However, the present study shows that early shortening and late prolongation of AV intervals significantly precedes the response of heart rate caused by the same autonomic perturbation (Fig. 2). The maximum shortening of AV interval preceded the maximum heart rate acceleration by 2–3 beats and maximum prolongation of AV interval preceded the maximum heart rate deceleration by approximately 4 beats. At the same time, the maximum prolongation of the AV interval was reached when the R-R interval was at the pre-VPB level (sixth post-VPB beat). During normal sinus rhythm, vagal stimulation prolongs the AV interval by slowing AV conduction. This effect is mitigated by concomitant slowing of heart rate, which facilitates AV conduction due to the recovery phenomenon. It has been shown that autonomic control of AV node conduction frequently dominates over the secondary “electrophysiologic” effect of heart rate change.3−6,10 Rarely and under specific circumstances, the effect of recovery may overwhelm the autonomic influences and result in an inverse relationship between R-R and AV intervals.1,10,17,18 In the present study, R-R and AV interval dynamics during HRT may seem to be positively associated indicating a predominance of autonomic modulation. However, the temporal dissociation of R-R and AV interval dynamics was evident. In addition, the phase of transfer function between R-R and AV intervals was near zero in an animal study during random pacing during autonomic blockade4 and similar observations were made in humans under normal autonomic control.10 Therefore, there are no experimental data to explain the main observation of our study. The early dynamics of AV conduction may be influenced by electrophysiological properties of the AV node. According to Jedlicka et al.,19 “it is a common observation that following interpolated or noninterpolated VPB the P-R interval is prolonged or the next P wave is dropped.” They hypothesized that this prolongation is due to AV nodal refractoriness secondary to retrograde
January 2003, Part II
PACE, Vol. 26
AUTONOMIC MODULATION OF AV CONDUCTION
conduction of the VPB. We actually observed just the opposite effect, that is a shortening of the first AV interval. Presumably, a longer interval between retrograde AV conduction and the first conducted sinus beat, which is obviously longer than the normal cycle length, considerably enhances the recovery of the AV node. Therefore, the early dynamics of AV intervals after a VPB may be more influenced by purely electrophysiological mechanisms rather than by vagal withdrawal. However, in the late phase, where nonautonomic effects are less likely to occur, the dissociation between R-R and AV interval response was even more pronounced. Since the fiducial points of QRS complexes and HRA potentials fell approximately in their centers, the measured AV intervals were practically equivalent to P-R intervals instead of the actual AV conduction times. However, since we studied their relative changes, this approximation was considered acceptable. The analytical approach (TO and TS descriptors) to quantify the dynam-
ics of R-R intervals after a VPB is not optimal for the description of a post-VPB pattern of AV intervals, which differs in the timing of early versus late responses. However, measuring the difference in amplitudes between post-VPB responses of R-R and AV intervals was not the main objective, since it was already known that the magnitude of autonomic modulation of AV conduction is significantly lower than that of R-R intervals. Conclusions Significant AV turbulence after VPBs was observed, consisting of an early shortening and late prolongation of AV intervals. Absolute magnitude of AV turbulence was much smaller than that of HRT. Therefore, dynamics of AV delay has a little impact on the accuracy of HRT assessment using the surface ECG. From a physiological point of view, the significant temporal dissociation of R-R and AV interval dynamics after a VPB remains unclear.
References 1. Martin P. Paradoxical dynamic interaction of heart period and vagal activity on atrioventricular conduction in the dog. Circ Res 1977; 40:81–9. 2. Wallick DW, Xu RG, Martin PJ. Dynamic interaction of vagal activity and heart rate on atrioventricular conduction. Am J Physiol 1992; 262:H792–H798. 3. O’Toole MF, Wurster RD, Phillips JG, et al. Parallel baroreceptor control of sinoatrial rate and atrioventricular conduction. Am J Physiol 1984; 246:H149–153. 4. Chen SL, Kawada T, Inagaki M, et al. Dynamic counterbalance between direct and indirect vagal controls of atrioventricular conduction in cats. Am J Physiol 1999; 277:H2129–2135. 5. Kawada T, Chen SL, Inagaki M, et al. Dynamic sympathetic control of atrioventricular conduction time and heart period. Am J Physiol 2001; 280:H1602–1607. 6. Mancia G, Bonazzi O, Pozzoni L, et al. Baroreceptor control of atrioventricular conduction in man. Circ Res 1979; 44:752–758. 7. Page RL, Tang AS, Prystowsky EN. Effect of continuous enhanced vagal tone on atrioventricular nodal and sinoatrial nodal function in humans. Circ Res 1991; 68:1614–1620. 8. Butrous GS, Cochrane T, Camm AJ. Rapid autonomic tone regulation of atrioventricular nodal conduction in man. Am Heart J 1987; 113:934–940. 9. Nollo G, Del Greco M, Ravelli F, et al. Evidence of low- and highfrequency oscillations in human AV interval variability: Evaluation with spectral analysis. Am J Physiol 1994; 267:H1410–1418. 10. Leffler CT, Saul JP, Cohen RJ. Rate-related and autonomic effects on atrioventricular conduction assessed through beat-to-beat PR
PACE, Vol. 26
11. 12.
13. 14. 15. 16. 17. 18.
19.
interval and cycle length variability. J Cardiovasc Electrophysiol 1994; 5:2–15. Schmidt G, Malik M, Barthel P, et al. Heart-rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial infarction. Lancet 1999; 353:1390–1396. Davies LC, Francis DP, Ponikowski P, et al. Relation of heart rate and blood pressure turbulence following premature ventricular complexes to baroreflex sensitivity in chronic congestive heart failure. Am J Cardiol 2001; 87:737–742. Mrowka R, Persson PB, Theres H, et al. Blunted arterial baroreflex cause “pathological” heart rate turbulence. Am J Physiol 2000; 279:R1171–R1175. Marine JE, Watanabe MA, Smith TW, et al. Effect of atropine on heart rate turbulence. Am J Cardiol 2002; 89:767–769. Grassi G, Seravalle G, Bertinieri G, et al. Sympathetic response to ventricular extrasystolic beats in hypertension and heart failure. Hypertension 2002; 39:886–891. Malik M, Wichterle D, Schmidt G. Heart rate turbulence. G Ital Cardiol 1999; 29(Suppl. 5):65–69. Donnerstein RL, Scott WA, Lloyd TR. Spontaneous beat-to-beat variations of PR interval in normal children. Am J Cardiol 1990; 66:753–754. Kowallik P, Meesmann M. Independent autonomic modulation of the human sinus and AV nodes: Evidence from beat-to-beat measurements of PR and PP intervals during sleep. J Cardiovasc Electrophysiol 1995; 6:993–1003. Jedlicka J, Martin P. Time course of vagal effects studied in clinical electrocardiograms. Eur Heart J 1987; 8:762–772.
January 2003, Part II
443
