Anesthetics and Automaticity of Dominant and Latent Pacemakers in Chronically Instrumented Dogs: IV. Dysrhythmias after Sinoatrial Node Excision

The methodology is identical to that reported previously, [14,15]with additions pertaining to the excision of the SA node (Figure 1). [12,13]After approval by the Institutional Animal Care Committee, a right thoracotomy was performed in eight mongrel dogs (18.5-21 kg), and the SA node along with adjacent atrial tissue (1.5 x 0.7 cm) was excised from the sulcus terminalis. [3,7,11-13]Epicardial electrodes were sutured to both atrial appendages and the ventricular apex. A patch, containing five electrode pairs, was sutured to the area on the right atrium beside the excision, extending to the junction with the inferior vena cava (Figure 2). A bipolar, interventricular needle electrode was used to record the His bundle electrogram. [16]Electrocardiogram (ECG) recordings during 3 weeks before testing verified that these dogs developed a stable subsidiary atrial pacemaker rhythm. The animals were allowed to recover from surgical placement of electrophysiologic monitors, during which time appropriate antibiotic and analgesic therapy was provided.

Three weeks after surgical implantation of monitoring devices, a 1-min paper record was obtained from the conscious, resting dog (control state), including the surface ECG and electrograms from the His bundle, the patch electrodes, and the left atrial appendage. Subsequently, conscious dogs were given intravenous infusions of 1 micro gram *symbol* kg sup -1 *symbol* min sup -1 epinephrine for 3 min. ECG and electrograms were recorded during the infusions. After 30 min, the epinephrine infusion was repeated at a rate of 2 micro gram *symbol* kg sup -1 *symbol* min sup -1. Anesthesia was induced via mask with inhalation of halothane, isoflurane, or enflurane in oxygen and was followed by tracheal intubation. Each dog was anesthetized with all three agents, with 5 days of rest between testing. The order of the anesthetics was randomly changed for each dog. End-tidal concentrations of the volatile agents and P^CO2were monitored continuously. Depth of anesthesia was maintained at 1.25 and finally 2 MAC. [17]End-tidal P^CO2was kept between 35 and 40 mmHg. After 30 min of equilibration, baseline recordings and data acquisition during epinephrine infusions were repeated. The above protocol was repeated under complete peripheral muscarinic blockade with intravenous methylatropine nitrate (3 mg *symbol* kg sup -1) administered 30 min before testing in the conscious state. The data were extracted from the same set of animals and experiments as reported previously. [12].

Each acquired heart beat was evaluated for the site of earliest identifiable myocardial activation (SEA) and assigned a score as follows (Figure 2): 1 = midrostral sulcus terminalis (3rd patch electrode); 2 = midcaudal sulcus terminalis (4th patch electrode); 3 = low caudal sulcus terminalis (5th patch electrode); 4 = caudal right atrial appendage area (2^ndpatch electrode); 5 = rostral right atrial appendage area (1^stpatch electrode); and 6 = remote atrial (left atrial appendage electrode). The highest remaining part of the sulcus terminalis contains the pacemaker with the highest automaticity and is labeled 1. Less active pacemakers are assigned increasing numeric values, as their automaticity declines within the atrium. Although high in the atrium, locations 4 and 5 are not on the sulcus terminalis and should be regarded as remote atrial locations. In this way, the SEA score reflects the functional dominance of the latent pacemakers, rather than merely the anatomic location. The fraction of beats from each origin was multiplied by the corresponding scores. Resulting products were summed and then divided by the fraction of beats of atrial origin, thus yielding an SEA value (range 1-6) independent of changes in rate and incidence of nonatrial beats. Because SEA values are not normally distributed data, nonparametric analysis of variances by Friedman ranks was used. Individual differences were assessed by the Wilcoxon's signed-rank test and the Kolmogorov-Smirnov test and represent effects of epinephrine (as compared to no epinephrine), effects of either anesthetic (compared to the corresponding conscious state condition), and effects of muscarinic blockade (compared to the absence of methylatropine nitrate). Differences between halothane, isoflurane, and enflurane also were analyzed. Nonatrial beats, such as His bundle beats (i.e., atrioventricular junctional), ventricular escape, and premature ventricular beats are counted separately, expressed as the summed percentages of all beats per condition (range 0-800), and compared in contingency tables by chi-square analyses. Significance was identified at P less or equal to 0.05. Data are shown as mean plus/minus SEM.

Immediately after SA node excision, qualitative changes in P wave morphology were observed on lead II of the surface ECG. Flat, biphasic, and inverted P waves were observed, often changing from one to another. Three weeks after SA node removal, the surface ECG showed predominantly upright P waves, and the midrostral portion of the sulcus terminalis (third patch electrode, score 1) was the most frequent SEA in conscious dogs without methylatropine or exposure to epinephrine (control state). Minimal phasic variations (< 25 ms) of heart rate with the respiratory cycle were present in conscious dogs and were accompanied by few pacemaker shifts to the midcaudal sulcus terminalis. These minor and infrequent shifts produced an SEA value of 1.183. Methylatropine completely abolished these variations, producing an SEA value of 1.0.

Without methylatropine, epinephrine induced shifts to lower pacemakers (Table 1), increasing SEA values in the conscious state, during halothane and isoflurane, but not enflurane anesthesia. However, with methylatropine, epinephrine increased SEA values in the conscious state and during 1.25 MAC halothane and 2 MAC enflurane but not isoflurane anesthesia. Methylatropine generally decreased the SEA value in the conscious state but increased it with 2 MAC enflurane and exposure to epinephrine. Without methylatropine, 2 MAC isoflurane decreased the SEA value during exposure to epinephrine 2 micro gram *symbol* kg sup -1 *symbol* min sup -1. With methylatropine, halothane increased the SEA value with epinephrine 1 micro gram *symbol* kg sup -1 *symbol* min sup -1. At 2 MAC with epinephrine, the SEA value with isoflurane was less than with halothane or enflurane.

Nonatrial dysrhythmias occurred only during exposure to epinephrine. The incidence of epinephrine-induced His bundle (i.e., atrioventricular junctional) beats was minimal in the conscious state and was increased by each anesthetic. Furthermore, His bundle beats were prevented by methylatropine during isoflurane anesthesia, whereas no such protection occurred during halothane or enflurane anesthesia. Epinephrine-induced ventricular escape beats occurred during severe bradycardia or when P waves were not observed on the surface or epicardial ECG. Ventricular escape beats were few in the conscious state, drastically increased with each volatile agent, and were always prevented by muscarinic blockade. Epinephrine-induced premature ventricular beats increased the most during halothane anesthesia, little during enflurane, but never during isoflurane anesthesia. Methylatropine had no appreciable effects on premature ventricular beats.

Excision of the SA node in chronically instrumented dogs was used as a model for sinus node dysfunction in patients to further characterize subsidiary atrial and lower pacemakers during exposure to anesthetics and epinephrine. [11-13]One limitation is that patients with clinical sinus node dysfunction may have variable degrees of disease-related changes within the region of the SA node. These pathologic processes also may affect subsidiary pacemakers, and therefore, these patients may have clinical arrhythmias not identified by this study. In addition, Rozanski et al. [9]showed less dependence of latent pacemakers on norepinephrine and greater sensitivity to acetylcholine than the SA node. The size of the SA nodal excision and the method employed (cryotherapy, infarct, excision) may determine the ability of autonomic nerves to regrow through the damaged area as well as the extent of denervation of the remaining pacemakers, because vagal fibers may pass through the excised area en route to lower pacemakers. Therefore, caution should be exercised when generalizing these data to other situations.

The SA node, normally the SEA, [14,15]usually over-drive-suppresses other potential pacemakers. [18]Previous reports on SA node excision [1-3,5,7-11,19,20]or destruction [3,4,6,21]described a period of bradycardia, atrial, and junctional escape rhythms, associated with flattened, inverted, or absent P waves on lead II of the surface ECG. A stable, regular, supraventricular rhythm, which developed several days after SA node excision, was characterized by an upright P wave, morphologically similar to that generated by SA rhythm. [8]During the testing for atrial pacemaker shifts in dogs with intact SA nodes, [14,15]only the shifts from the SA node to the caudal sulcus terminalis were accompanied by distinctly altered P waves. Changes of P wave morphology in this study are consistent with those reported by other authors [8]and our previous studies. [14,15]Considering that P waves from intact dogs have considerable morphologic variability, [14,15]we conclude that SA rhythms cannot be distinguished from chronic subsidiary atrial pacemaker rhythms from the limb leads of the surface ECG.

Along the sulcus terminalis, spontaneous pacemaker automaticity generally decreases with increasing distance from the SA node. [14,15]In conscious, resting dogs, subsidiary atrial pacemakers near the SA node have similar automaticity to the SA node, [3,8,12,13,20]although the subsidiary atrial pacemakers operate closer to their maximum rate. [12-15]Pacemaker shifts occur whenever the automaticity of the primary pacemaker is sufficiently reduced below that of subsidiary pacemakers, [14,15]sometimes producing clinically significant rhythm disturbances. [22]Baroreflex-mediated increases of vagal tone are one mechanism responsible for pace-maker shifts. [14,15]Selective decrease of SA node automaticity may result because greater density of vagal innervation exists in the SA node than in other atrial sites, [3,23]thereby subjecting the SA node to greater vagal pacemaker suppression. Greater muscarinic receptor density may render the SA node more susceptible to vagal suppression, [9,24]and differences in intrinsic discharge rates may exist between the SA node and subsidiary pacemakers, possibly due to differences in ion channel distribution, pacemaker currents and action potential characteristics. [25-28].

Because the current experimental protocol is identical to that previously published on dogs with intact SA nodes, [14,15]comparisons between these groups can be made, and summarized data from dogs with intact SA nodes are listed in Table 1. In the control state, the primary pacemaker in intact dogs was most frequently the SA node, which was given a score of 1, whereas the midrostral sulcus terminalis was given a score of 3. [14,15]In the control state after SA node excision, the primary cardiac pacemaker is located at the midrostral sulcus terminalis and is assigned a score of 1 (Figure 2). This makes SEA values functional rather than anatomic measures, assigning the lowest score to pacemakers with the highest automaticity under control conditions.

Dogs without SA nodes had greater SEA values than did intact dogs during halothane anesthesia without methylatropine under some experimental conditions but lower SEA values during 2 MAC isoflurane anesthesia. In the conscious state, during halothane or isoflurane anesthesia, and with methylatropine in the absence of epinephrine, dogs without SA nodes had lower SEA values than did intact dogs. During 1.25 MAC halothane and epinephrine 1 micro gram *symbol* kg sup -1 *symbol* min sup -1, the SEA value was increased compared to intact dogs. [14,15].

His bundle beats occurred more frequently in dogs without SA nodes during halothane and enflurane anesthesia, regardless of muscarinic blockade. Without methylatropine during 1.25 MAC isoflurane, the incidence of His bundle beats was zero, less than dogs with intact SA nodes. [14,15]However, at 2 MAC isoflurane, the incidence was increased. [13]Compared to dogs with intact SA nodes, no increase of His bundle beats or ventricular escape beats occurred in the conscious state. [13]However, during each anesthetic tested, the incidence of ventricular escape beats was much greater after SA node excision. [13-15]Finally, in the conscious state and during halothane or enflurane anesthesia, dogs with excised SA nodes had a greater incidence of premature ventricular beats than dogs with intact SA nodes. [14,15].

The existence of pacemaker shifts among subsidiary atrial sites after SA node excision suggests that important differences in automaticity, vagal innervation, and/or intrinsic pacemaker properties also exist among subsidiary atrial pacemakers after SA node excision. Our findings indicate that muscarinic transmission to the heart is required for variations of cycle lengths with respiration. These variations are much smaller compared to the respiratory sinus dysrhythmia in dogs with intact SA nodes, [14,15]which supports the contention that less vagal innervation exists in the midrostral sulcus terminalis. In previous studies, where muscarinic transmission was blocked, epinephrine resulted in an increase in the heart rate, but with intact muscarinic transmissions, heart rate generally decreased when epinephrine was administered. [12,14,15]Furthermore, in this study, subsidiary atrial pacemakers were suppressed by increased vagal tone in a progressive and hierarchical fashion, and muscarinic transmission was necessary for pacemaker shifts from midrostral to caudal atrial sites in conscious and isoflurane-anesthetized dogs. [12,13]Finally, ventricular escape beats only develop during bradycardia when muscarinic transmission to the heart is intact, regardless of the test condition or the presence [14,15]or absence of the SA node. [12,13].

Volatile anesthetics can exert direct negative chrono-and dromotropic effects on the SA node and the cardiac conduction system, [12-15,29-33]but the degree to which this occurs in subsidiary atrial pacemakers has not been determined. Compared to the conscious state and to dogs with intact SA nodes, [14,15]dogs with excised SA nodes are more likely to develop epinephrine-induced ventricular escape during anesthesia, particularly with isoflurane. [12,13]Intact SA node function protects against ventricular escape rhythms during anesthesia, and the hierarchical sequence of vagal pacemaker suppression can result in complete atrial standstill and the emergence of a ventricular pacemaker. [13]Isoflurane is likely to impair the baroreflex the least, allowing for the greatest vagally mediated atrial pacemaker suppression, thereby causing the greatest incidence of ventricular escape. [12,13]Ventricular escape rhythms with isoflurane-and epinephrine-increased blood pressure are preceded by severe bradycardia and atrial asystole. During halothane anesthesia, bradycardia is less prominent, whereas with enflurane, supraventricular rate-changes barely occur. [12,13].

Both overdrive and underdrive suppression may be impaired in the presence of chronic subsidiary atrial pacemaker rhythm. Dominant pacemakers overdrive-suppress subsidiary pacemakers by depolarizing at a greater rate, triggering their depolarization and preventing them from becoming the sites of earliest activation. [34]Overdrive suppression can fail when the spontaneous rate of the subsidiary pacemaker becomes greater than the rate of the dominant pacemaker. Underdrive suppression occurs when the dominant pacemaker prevents the escape of subsidiary pacemakers, even though the spontaneous rate of the dominant pacemaker is less than that of the subsidiary pacemaker after escape. [4]Although equal direct anesthetic depression and little vagal suppressibility may exist in ventricular pacemakers, anesthetics may suppress subsidiary atrial pacemakers more than the SA node. Prolonged recovery from overdrive-pacing exists in subsidiary atrial pacemakers, [14,15]and when exposed to epinephrine, their rates in isoflurane-[13]and enflurane-anesthetized dogs are less than SA rates, [14,15]supporting the interpretation that subsidiary atrial pacemakers have less overdrive capability. However, because subsidiary atrial pacemakers adapt over time to the absence of the SA node, [7,9]ventricular pacemakers may undergo a positive chronotropic adaptation to the slower trigger rates by subsidiary atrial pacemakers, possibly contributing to the increase of escape beats after SA node excision. Current findings suggest that subsidiary atrial pacemakers are less able to underdrive- or overdrive-suppress ventricular pacemakers than the SA node. [12,13].

Volatile anesthetics can influence cardiac dysrhythmias through autonomic, humoral, and cellular mechanisms. [31,35-41]Sensitization is an interaction between catecholamines and volatile anesthetics, which is most commonly associated with lowering the threshold for ventricular dysrhythmias, [36-39,42-45]but investigations also determined the dose of epinephrine facilitating wandering atrial pacemaker, atrial ectopies, ventricular escape, and atrioventricular junctional rhythms. [42,46]In anesthetized dogs with intact SA nodes, pacemaker shifts that occurred in the presence of methylatropine were accompanied by increases of pacemaker automaticity, suggesting that epinephrine-anesthetic interaction may enhance the automaticity more in subsidiary pacemakers than in the SA node. [14,15]In the current study, epinephrine-induced shifts from the midrostral to remote atrial pacemaker locations are exacerbated by halothane, providing evidence that subsidiary atrial pacemakers are susceptible to epinephrine-anesthetic sensitization, consistent with greater enhancement of automaticity in remote atrial than in midrostral atrial pacemakers. [12]In contrast, neither isoflurane nor enflurane sensitize the subsidiary atrial pacemakers in the current study, as shown by smaller SEA values in comparison to halothane. Pacemaker shifts with halothane and epinephrine in this study exceed those from studies with intact SA nodes, [14,15]allowing the conclusion that intact SA node function protects against supraventricular dysrhythmias during halothane anesthesia. In contrast, antidysrhythmic actions of volatile agents [31,47,48]may be due in part to reductions in autonomic tone. [49-53]Although reduced sympathetic tone may decrease ventricular tachydysrhythmias, anesthetics may attenuate the baroreflex and decrease parasympathetic tone. In our study, 2 MAC isoflurane desensitizes subsidiary atrial pacemakers, and SA node excision appears to reduce epinephrine-induced pacemaker shifts with isoflurane. It is likely that isoflurane has little direct dysrhythmogenic effects on subsidiary atrial pacemakers, and the reduction of atrial pacemaker shifts is due to decreased autonomic tone and a direct and equally depressant effect of isoflurane on all subsidiary atrial pacemakers. [13].

Results from this study parallel previous findings [14,15]that halothane and enflurane sensitize the heart to His bundle dysrhythmias and premature ventricular beats. Intact SA node function [14,15]appears protective against His bundle dysrhythmias and premature ventricular beats. More junctional and ventricular dysrhythmias can be expected, secondary to direct effects of epinephrine or to epinephrine-anesthetic interaction. Our data suggest this interaction for enflurane, because infused epinephrine does not compensate for negative chronotropic effects of enflurane, [12]and His bundle and ventricular dysrhythmias are more frequent without the SA node. However, other mechanisms, such as the effects of heart rate, cardiac output, plasma epinephrine levels, blood pressure, and ventricular stretch on ventricular dysrhythmias cannot be excluded on the basis of this study.

Summarizing findings from chronically instrumented dogs after SA node excision, we conclude that: (1) subsidiary atrial pacemakers show less vagal innervation than SA node pacemakers; (2) after SA node excision, severe supraventricular bradycardia and ventricular escape result from epinephrine-anesthetic interactions, especially with isoflurane anesthesia; (3) SA node excision generally facilitates ventricular escape, His bundle dysrhythmias, and premature ventricular beats; and (4) adaptation processes in ventricular pacemakers may occur after SA node excision and need to be investigated further. If these findings can be extrapolated to humans considering the above mentioned limitations, they suggest that patients with SA node dysfunction may be more susceptible to severe cardiac dysrhythmias during epinephrine-anesthetic interactions, even if SA node dysfunction is not evident in the conscious state.

The authors thank Dr. R. Hoffman, for his valuable advice in statistical methods, and J. Krolikowski, for his technical assistance.