Modulation of Atrial Repolarization by Site of Pacing in the Isolated Rabbit Heart
Background Single-site or multisite atrial pacing may reduce the incidence of atrial fibrillation in humans. The therapeutic mechanisms may include synchronization of atrial repolarization (repolarization “memory”) and/or decreased dispersion of atrial repolarization. These responses have not been well documented in intact atria.
Methods and Results Monophasic action potential recordings were made from six atrial epicardial sites in 39 isolated perfused rabbit heart preparations during 3 hours of continuous right atrial, left atrial, or biatrial pacing. Action potential recordings obtained at times 0, 45, 90, 135, and 180 minutes were computer analyzed for activation time (AT) and 90% action potential duration (APD) at each site. No consistent relationship could be demonstrated between APD and AT at any time during atrial pacing (all P>.05). On average, left atrial APDs were longer than right atrial APDs by up to 6.3 ms at all times, regardless of the site of pacing (P≤.05). At all times, dispersion of atrial repolarization was minimized by left atrial pacing compared with right atrial pacing (21.6±9.1 versus 32.4±15.1 ms, respectively, at time 0; P<.05). Biatrial pacing provided no further reduction in dispersion of repolarization compared with left atrial pacing (all P>.05).
Conclusions No relationship can be demonstrated between atrial AT and APD in the isolated rabbit heart preparation. This differs from ventricular repolarization “memory,” which is demonstrable under the same conditions. Left atrial APD is, on average, longer than right atrial APD, suggesting spatial heterogeneity in repolarization. Dispersion of atrial repolarization is minimized by left atrial pacing in this preparation with no further advantage to biatrial pacing.
The pharmacological therapy of atrial fibrillation is frequently ineffective, poorly tolerated, or dangerous because of proarrhythmic responses.1 2 3 4 These shortcomings have led to the investigation of nonpharmacological therapies, such as surgery, catheter ablation, and permanent cardiac pacing, as primary or adjunctive antiarrhythmic therapies.5 6 7 8 9 Retrospective data suggest that AV sequential pacing may reduce the incidence of atrial fibrillation in patients with sick sinus syndrome, although the therapeutic mechanism is unknown.6 Although improved hemodynamics may be beneficial, the primary electrical effects of atrial pacing also may be therapeutically important. If the timing and sequence of atrial activation can be regularized, the dispersion of atrial repolarization may be reduced.10 11 Simultaneous multisite atrial pacing has been proposed to further synchronize atrial activation and repolarization, although beneficial electrophysiological effects have not been clearly demonstrated.12 13
Ventricular myocardium exhibits the property of repolarization “memory.” Alteration of the ventricular depolarization sequence modulates the repolarization sequence to reestablish orderly and synchronized repolarization.14 15 16 17 By synchronizing global ventricular repolarization, this process may provide intrinsic antiarrhythmic effects.18 It is unknown whether this antiarrhythmic property exists in the atrium or can be invoked by artificial pacing of atrial tissue. If it is present, judicious atrial pacing protocols may offer a new option for antiarrhythmic therapy.
The purpose of this study was to characterize the influence of the site of atrial stimulation on atrial repolarization in an intact isolated rabbit heart preparation. Specific study objectives were to (1) determine the relationship between regional atrial activation sequence and APD, (2) identify factors that influence the regional atrial APD, and (3) determine the influence of stimulation site on the dispersion of atrial repolarization.
Isolated Heart Preparation
New Zealand White rabbits of either sex weighing 3.5 to 4.5 kg were fully anesthetized with ketamine (35 mg/kg IM) and xylazine (7 mg/kg IM) in accordance with the American Veterinary Medical Association Panel on Euthanasia guidelines. After administration of 500 IU heparin IV, median sternotomies and cardiectomies were rapidly performed. The aorta was immediately secured to a cannula while being perfused manually with oxygenated perfusate solution, then rapidly transferred to a vertical Langendorff apparatus and perfused with oxygenated (95% O2/5% CO2) Krebs-Henseleit solution (in mmol/L: NaCl 118, CaCl2 2.5, KCl 4.7, MgSO4 1.2, KH2PO4 1.2, NaHCO3 25, glucose 5.5) with 0.0006 mmol/L bovine albumin added to enhance the electrophysiological stability of the preparation.14 19 The perfusate was maintained at pH 7.35 to 7.45, Po2 >500 mm Hg, and Pco2 28 to 35 mm Hg throughout the experiment. A constant perfusion pressure of 75 mm Hg was maintained by a fluid column. The left ventricle was vented with a 20-gauge needle, and all extraneous tissue was dissected from the atrial surfaces. The entire heart was submerged in a fluid bath with perfusate and bath temperature maintained at 37.0°C (Yellow Springs Instruments Precision 4000; 0.02°C resolution).
Two Teflon-coated silver wire pacing electrodes were affixed to the tip of either the right, left, or both atrial appendages. Constant-current atrial pacing was used, with a 250-ms cycle length and a 2.0-ms rectangular pulse at twice diastolic threshold (Bloom Associates DTU 201).
Atrial MAP recordings were made by use of custom-built suction electrode probes applied to the epicardial atrial surface. Suction electrodes obtain stable recordings on the intact beating atria, and these recordings accurately reflect RTs.20 The probes comprised a hollow plastic cylinder (ID, 1 mm; OD, 2 mm) with 0.25-mm-diameter insulated silver wire electrodes set 0.5 mm from the probe tip on the inner lumen and external walls. The probes were positioned with micromanipulators, and vacuum (−23 psi) was applied to the lumen to initiate MAP recording. Recordings were suitable for analysis if, after 15 to 20 seconds of stabilization, the peak amplitude was >9 mV, dV/dt of phase 0 was >9 V/s, and a triangular or slightly plateaued MAP was recorded.21 22 23
Data Acquisition and Analysis
The MAPs were amplified and band-pass filtered at 0.01 to 1000 Hz with DC coupled amplifiers (World Precision Instruments, ISO-DAM 8). Beginning with the pacing impulse, the MAP was digitally sampled (16 bits) at 4000 Hz for 200 ms under the control of custom software written in ASYST (Keithley Instruments). At each atrial recording site, 50 consecutive MAPs were sampled, stored, and averaged to generate a single composite tracing. The averaged MAP from each site underwent automated analysis to measure (1) AT, the time interval from the pacing artifact to onset of phase 0; (2) APD, the time interval from phase 0 onset to 90% repolarization; (3) peak action potential amplitude; and (4) maximum dV/dt of phase 0. The RT for each site was defined as the sum of AT plus APD for that site14 (Fig 1⇓). The dispersion of atrial RTs was defined as the difference between the longest and shortest values of the six RTs recorded from each heart at each sampling time.
Atrial pacing was begun within 10 minutes of initiation of Langendorff perfusion and was maintained without interruption for the entire experiment. Each heart underwent either continuous right (n=13), left (n=13), or biatrial (n=13) pacing to define three study groups based on the site(s) of pacing. At time 0, defined as 5 minutes after the start of atrial pacing, MAP recordings were made from six epicardial atrial sites defined as follows: site 1, as close as possible to the pacing electrodes at the tip of the atrial appendage; site 2, the center of the atrial free wall midway between the paced atrial appendage tip and the ipsilateral atrial septum; site 3, the ipsilateral atrial midseptal border of the paced atrium; site 4, the contralateral (unpaced) atrial midseptal border; site 5, the center of the atrial free wall midway between the contralateral (unpaced) septal border and the atrial appendage tip; and site 6, the tip of the contralateral (unpaced) atrial appendage (Fig 2⇓). For biatrial pacing, the six recording sites were classified as sites 1 through 3 as defined above in both the right and left atria. All six recordings were made within a 3- to 5-minute interval with two probes. MAP recordings from the six sites were repeated after 45, 90, 135, and 180 minutes of continuous pacing. The recordings were made as close as possible to the initial (time 0) sites in each heart; however, repeated recordings from identical sites often were not possible, probably because of focal areas of suction-induced trauma from the probes. In total, MAP recordings were made from six sites at each of five times in 39 hearts. The 180-minute recording duration was chosen to give adequate time for the development of new repolarization patterns while still recording stable MAPs. Experiments that failed to produce complete recordings at all time intervals were discarded. Perfusate pH, Po2, Pco2, and bath temperature were checked just before each set of recordings.
Statistical analysis was performed with commercially available software (SPSS for Windows 6.0). All values are expressed as mean±SD. Repeated-measures ANOVA was used to assess the stability in study variables over time. Analysis for correlations between AT and APD was performed by linear regression analysis. For a relationship with an R2=.25, with 39 subjects, significance was detected with a power of 0.95. ANOVA was used to analyze the association of APD with noncontinuous variables. The differences in dispersion of RTs between pacing groups were analyzed by one-way ANOVA with least-significant-difference post hoc analysis. Student's t test was used for paired comparisons between independent groups. A value of P<.05 was considered significant.
The perfusate pH and oxygenation remained within the defined limits for all experiments. The average coronary perfusate flow rate was 59±21 mL/min. To assess the stability of the repeated measurements over time, the values for MAP amplitude, APD, and AT from all sites in all 39 hearts were pooled and compared across the five recording times. Maximal MAP amplitude averaged 22.3±4.2 mV at time 0 and showed no significant change across the five recording times (P=.13). Mean AT varied by only 1.8 ms across the five recording times (P=.016). The average value of APD decreased linearly with time, decreasing by 9.2 ms from time 0 to 180 minutes (87.4±8.0 to 78.2±5.6 ms, respectively, P<.001). A linear reduction in APD occurred in both the right and left atria. Serial reductions in APD have been described previously for analogous ventricular MAP recordings.14 Such changes did not obscure the relationship between ventricular AT and APD and repolarization memory, however.14
As expected, AT increased with increasing distance from the pacing site. AT at sites 1 through 6 averaged 9.1±5.5, 15.2±5.4, 17.7±5.0, 28.2±5.6, 30.5±4.7, and 30.0±4.2 ms, respectively, at time 0. The average value of AT from any recording site varied by no more than 2.9 ms across the five recording times. The average time for atrial activation was significantly influenced by single or biatrial pacing sites at all times. The average value for AT from all six sites at time 0 was 21.5±10.1 ms for hearts undergoing right atrial pacing, 21.4±9.4 ms for left atrial pacing (P>.05 versus right), and 14.7±7.0 ms for biatrial pacing (P<.001 versus right or left pacing).
Relationship Between AT and APD
If repolarization memory exists in atrial tissue, then an inverse relationship between AT and APD should be demonstrated for each site of atrial pacing. This relationship was sought by three statistical approaches. First, regression analysis performed for each of the 39 hearts at each of the five sample times failed to demonstrate consistent and statistically meaningful relationships between AT and APD. At 45 minutes, 9 of 39 hearts demonstrated a significant relation between AT and APD (all P≤.04), a proportion no better than chance. At times 0, 90, 135, and 180 minutes, only 7, 4, 4, and 7 hearts, respectively, demonstrated significant relationships between AT and APD, again proportions no different from chance. In addition, these relationships were inconsistent in the direction of the regression coefficients between individuals. These relations were direct (B was positive) in 13 hearts (9 right and 4 biatrially paced) and indirect (B was negative) in 18 (9 left, 7 right, and 2 biatrially paced). To ensure that a relationship between AT and APD was not masked by undersampling of data points, six additional hearts underwent continuous right (n=3) or left (n=3) atrial pacing with 20 atrial sites (10 right, 10 left) recorded at times 0 and 180 minutes. Regression analysis was performed for each heart. At time 0, only 2 hearts showed significant relationships between AT and APD. Both hearts were left atrially paced, and the relationship between AT and APD was direct for one and inverse for the other (both P≤.006). At 180 minutes, 2 hearts also had significant relationships between AT and APD. In 1 left atrially paced heart, the relationship was inverse, and in one right atrially paced heart, the relationship was direct between AT and APD (both P≤.049). These variable and inconsistent relationships between AT and APD may be explained by the intrinsically longer APD in the left atrium compared with the right (see below).
Second, regression analysis was performed with all data points from each heart at each time pooled to evaluate for relations between AT and APD. No significant relationship between absolute or standardized values of AT and APD could be demonstrated at any time (all P≥.089). AT and APD values were standardized for each heart at each time by the usual method, using the mean and SD of the six individual values at each time. This analysis was used to minimize the between-subject absolute variations in AT and APD in the group analysis. Similarly, no relationship between AT and APD could be found within the right or left atrium at any of the five times when right and left atrial sites were analyzed separately (all P≥.16).
Third, the relationship between APD and activation sequence was also examined, comparing values for APD with recording sites 1 through 6 in the 26 hearts undergoing right or left atrial pacing. Hearts undergoing biatrial pacing were excluded from this analysis because recordings from an unpaced atrium were not available. The site of recording as a representation of AT and anatomic distance from the pacing site had no significant relation to APD at any of the five recording times (all P≥.10). In summary, no consistent relationship could be demonstrated between APD and either AT or anatomic distance from pacing site by any analysis.
Determinants of APD
Given the inability of AT to account for variability in APD, the influence of right versus left atrial recording site and site of pacing on APD was examined. The ANOVA model demonstrated left atrial sites to be associated with longer APDs than right atrial sites at all times except 45 minutes (all P≤.04 except 45 minutes, P=.23). At time 0, the average left and right atrial APDs were 88.9±10.6 and 85.8±9.6 ms, respectively (P=.035). Left atrial APD averaged 1.6 to 6.3 ms longer than right atrial APD at all subsequent times (difference, 1.6 ms at time 45 minutes). Site of pacing (right, left, or biatrial) was not associated with APD at any time (all P>.05).
To assess for anisotropic influences on APD from the sequence of activation, the values of APD from the same anatomic sites were compared at each time between hearts undergoing right and left atrial pacing. For recordings from the same anatomic site, APD was significantly different between left and right atrially paced hearts only at the left atrial appendage site (corresponding to site 6 in Fig 2⇑) at time 0 (right atrial pacing, 87.4±13.5 ms; left atrial pacing, 78.5±7.04 ms; P=.025) and the left atrial paraseptal site (corresponding to site 4 in Fig 2⇑) at time 3 (right pacing, 84.3±7.01 ms; left pacing, 81.9±11.4 ms; P=.006).
Dispersion of Atrial Repolarization
The dispersion of atrial repolarization was significantly influenced by the site of pacing at all five recording times (Fig 3⇓). At time 0, dispersion of repolarization was 32.4±15.1, 21.6±9.1, and 23.0±6.2 ms for right, left, and biatrial pacing, respectively. Dispersion of repolarization was significantly smaller for left atrial pacing than for right atrial pacing at all times (all P<.05) (Fig 4⇓). Dispersion of repolarization was similar for left and biatrial pacing at all times (all P>.05). Dispersion of repolarization was greater for right atrial than for biatrial pacing at times 0, 45, and 180 minutes (all P<.05). The average value of dispersion of atrial repolarization did not change over time for any of the three pacing groups (all P>.05).
Three major findings result from the present study. First, no consistent relationship between atrial AT and APD could be demonstrated during atrial pacing in the isolated rabbit heart. This result occurred despite study of 39 hearts under conditions that allow detection of ventricular memory,14 despite extensive mapping (20 sites) in some hearts, and despite analysis of individual and pooled data. The statistical methodology had high power (0.95) and in fact was biased toward the demonstration of AT-APD relationships by use of relatively few sample sites (6) per heart, use of pooled data, and performance of multiple analyses. Second, left atrial APD is longer, on average, than right atrial APD, suggesting intrinsic spatial differences in repolarization properties. Third, although average atrial AT was reduced by biatrial pacing, the dispersion of repolarization is minimized by left atrial pacing alone in this preparation.
The absence of a consistent relationship between AT and APD independent of the site of pacing in atrial tissue differs from findings in ventricular myocardium. Costard-Jackle et al14 demonstrated an inverse relationship between ventricular AT and APD in isolated rabbit hearts during sinus rhythm. This inverse relationship disappeared within 5 minutes of ventricular pacing but was reestablished within 60 minutes of continued pacing, ie, APD conformed to the new activation sequence rather than being a fixed property dependent on location. This inverse ventricular AT-APD relationship has been documented in humans during open heart surgery, and time- and rate-dependent changes in T-wave morphology after ventricular pacing are also consistent with altered patterns of repolarization.15 16 17 In our study, the discordant and inconsistent relationships between AT and APD in the right and left atrially paced hearts can be explained by the intrinsically longer left atrial APD. With right atrial pacing, these longer left atrial APDs are associated with longer ATs, giving a direct relationship. With left atrial pacing, the longer left atrial APDs were associated with shorter ATs, giving an inverse relationship. Because the average differences in APD between right and left atria are small, these relationships between APD and AT did not always rise to the level of significance.
The absence of repolarization memory in atrial tissue may result from many factors. In ventricular myocardium, the memory phenomenon may be due in part to electrotonic interactions by which electromotive forces between muscle fibers in different states of activation influence the subsequent sequence of repolarization.16 24 The distribution and flow of such electrotonic currents are highly dependent on membrane currents, intercellular junctions, tissue mass, tissue geometry, and fiber orientation. Histologically, atrial myocardial fiber orientation is disorganized compared with ventricular tissue, and the anatomy of both atria is highly irregular.25 In ventricular tissue, repolarization memory is abolished by Ito blockade with 4-aminopyridine, possibly as a result of the loss of the transmural gradient in the magnitude of this current.26 Although Ito is critical to repolarization of atrial myocytes, the reduced wall thickness and mass of the atrium may preclude the memory phenomenon. Differences in the contribution of other currents to setting APD, the higher membrane resistance of atrial versus ventricular tissue, and differences in the quantity and structure of intercellular gap junctions may also prevent the memory phenomenon.27 28 Our findings are consistent with those of Spach et al,11 who demonstrated highly variable shapes and durations of action potentials within isolated rabbit right atrial tissue. These variations in APD appeared to show clear spatial organization but were not influenced by sequence of activation, loss of electrotonic influences, or muscarinic blockade. APD did correspond to regional expression of troponin isoforms, however, suggesting a genetic regulation of repolarization.
In our study, left atrial APD was, on average, longer than right atrial APD regardless of the site of pacing. The reasons for this difference are unclear but may be a further extension of the regional intrinsic differences in APD demonstrated by Spach et al.10 From these studies, it appears that atrial APD varies between as well as within the right and left atria. Although this finding may represent intrinsic behavior of the myocytes, APD also may be modulated by the pressure differential between the atria or differential autonomic innervation in vivo.
The disparity between right and left atrial APD becomes important when the dispersion of atrial repolarization is analyzed. Given the tendency for longer left atrial APD compared with right atrial APD, the differences in RTs between the atria in this preparation are determined largely by the ATs. If the left atrium is activated late during right atrial pacing, the dispersion of repolarization is maximized, because longer left atrial APDs are coupled with longer ATs and shorter right atrial APDs are associated with shorter ATs. Conversely, the dispersion of repolarization is minimized by left atrial pacing. During simultaneous biatrial activation, right and left ATs are comparable, but the net effect of longer left atrial APD on dispersion of repolarization persists. There was no site-specific anisotropic influence of activation sequence on APD. This finding suggests a benefit from left atrial pacing but raises the possibility that biatrial pacing may provide little additional advantage. This finding requires verification in the hemodynamically loaded and innervated heart, however.
The isolated heart preparation is devoid of autonomic and hemodynamic influences. Although these factors have not been implicated in ventricular memory, the possibility that they are necessary for atrial memory cannot be excluded. In addition, these same factors could equalize or enhance the disparities between left and right atrial APD noted in this study. As is common for isolated tissue preparations, APD decreased over time. Similar global APD shortening does not mask the memory phenomenon in ventricular tissue, however.14 True transmembrane action potentials were not recorded in this study; however, suction electrodes have been shown to accurately reflect the time course of myocardial repolarization.20
The absence of atrial repolarization memory and the fixed spatial differences in atrial APD may contribute to atrial vulnerability to fibrillation. This vulnerability may in part arise from the inability of the atrium to alter increased dispersion of refractoriness that results from unfavorable patterns of activation. Despite the tendency to prolong the dispersion of atrial repolarization, the clinical benefit from right atrial pacing reported in prevention of atrial fibrillation may arise from the prevention of atrial pauses, and even further benefit may be expected from the addition of left atrial pacing. The reduction of atrial dispersion of refractoriness by left atrial pacing in this study may explain the electrophysiological benefit of biatrial pacing reported in preliminary clinical work.12 13 This finding also supports the possibility that complex biatrial lead systems may be unnecessary if the majority of clinical benefit derives from left atrial pacing alone.
Selected Abbreviations and Acronyms
|APD||=||action potential duration|
|MAP||=||monophasic action potential|
The authors thank Henry Clemo, MD, PhD, for his helpful comments and technical advice; Ann Sykes, BS, for her expert technical work; and Kay Lentz for preparation of this manuscript.
- Received January 8, 1996.
- Revision received March 13, 1996.
- Accepted April 7, 1996.
- Copyright © 1996 by American Heart Association
Coplen SE, Antman EM, Berlin JA, Hewit P, Chalmers TC. Efficacy and safety of quinidine therapy for maintenance of sinus rhythm after cardioversion. Circulation. 1990;82:1106-1116.
Harris L, McKenna WJ, Rowland E, Holt DW, Storey GCA, Krikler DM. Side effects of long-term amiodarone therapy. Circulation. 1983;67:45-51.
Velebit V, Podrid P, Lown B, Cohen BH, Graboys TB. Aggravation and provocation of ventricular arrhythmias by antiarrhythmic drugs. Circulation. 1982;65:886-894.
Ayers GM, Alferness CA, Ilina M, Wagner DO, Sirokman WA, Adams JM, Griffin JC. Ventricular proarrhythmic effects of ventricular cycle length and shock strength in a sheep model of transvenous atrial defibrillation. Circulation. 1994;89:413-422.
Swartz JF, Pellersels G, Silvers J, Patten L, Cervantez. A catheter-based curative approach to atrial fibrillation in humans. Circulation. 1994;90(suppl I):I-335. Abstract.
Spach MS, Dolber PC, Anderson PAW. Multiple regional differences in cellular properties that regulate repolarization and contraction in the right atrium of adult and newborn dogs. Circ Res. 1989;65:1594-1611
Spach MS, Dolber PC, Heidlage JF. Interaction of inhomogeneities of repolarization with anisotropic propagation in dog atria. Circ Res. 1989;65:1612-1631.
Costard-Jackle A, Goetsch B, Antz M, Franz M. Slow and long-lasting modulation of myocardial repolarization produced by ectopic activation in isolated rabbit hearts. Circulation. 1989;80:1412-1420.
Franz MR, Bargheer K, Rafflenbeul W, Haverich A, Lichtlen PR. Monophasic action potential mapping in human subjects with normal electrocardiograms: direct evidence for the genesis of the T wave. Circulation. 1987;75:379-386.
Kuo C-S, Munakata K, Reddy CP, Surawicz B. Characteristics and possible mechanisms of ventricular arrhythmia dependent on the dispersion of action potential durations. Circulation. 1983;67:1356-1367.
Hoffman BF, Cranefield PF, Lepeschkin E, Surawicz B, Herrlich HC. Comparison of cardiac monophasic action potentials recorded by intracellular and suction electrodes. Am J Physiol. 1959;196:1297-1301.
De Carvalho AP, De Mello WC, Hoffman BF. Electrophysiologic evidence for specialized fiber types in rabbit atrium. Am J Physiol. 1959;196:483-488.
Gelband H, Bush HL, Rosen MR, Myerburg RJ, Hoffman BF. Electrophysiologic properties of isolated preparations of human atrial myocardium. Circ Res. 1972;30:293-300.
Hogan PM, Davis LD. Evidence for specialized fibers in the canine right atrium. Circ Res. 1968;23:387-396.
Joyner RW. Modulation of repolarization by electrotonic interactions. Jpn Heart J. 1986;27:167-183.
Bharati S, Lev M. Histology of the normal and diseased atrium. In: Falk RH, Podrid PJ, eds. Atrial Fibrillation: Mechanisms and Management. New York, NY: Raven Press Ltd; 1992:15-39.
Del Balzo V, Rosen MR. T wave changes persisting after ventricular pacing in canine hearts are altered by 4-aminopyridine but not by lidocaine. Circulation. 1992;85:1464-1472.