Management of patients with sinus node dysfunction must consider the stability of subsidiary pacemakers during anesthesia and treatment with antimuscarinic or sympathomimetic drugs. Baroreflex regulation of atrial pacemaker function is known to contribute to the interactions between inhalation anesthetics and catecholamines. Sinoatrial (SA) node excision can be a model for intrinsic SA node dysfunction. Subsidiary atrial pacemakers are expected to emerge after SA node excision, but they may respond differently to humoral and neural modulation. Isolated and combined effects of epinephrine and methylatropine should help characterize subsidiary pacemaker function during anesthesia with halothane, isoflurane, and enflurane.
In eight dogs, SA nodes were excised and epicardial electrodes implanted at the atrial appendages, the His bundle, and along the sulcus terminalis. Spontaneous pacemaker automaticity and subsidiary atrial pacemaker recovery time were measured in the conscious state, in the presence of methylatropine, with 1 and 2 micrograms.kg-1.min-1 epinephrine and during 1.25 and 2 MAC halothane, isoflurane, and enflurane.
After SA node excision, a stable and regular subsidiary atrial pacemaker rhythm emerged. Each anesthetic prolonged subsidiary atrial pacemaker recovery times. This prolongation was greater in the presence of methylatropine. Without methylatropine, isoflurane and enflurane, but not halothane, further enhanced the baroreflex-mediated negative chronotropic effects of epinephrine, whereas with methylatropine, each anesthetic reduced the direct positive chronotropic effects of epinephrine.
Halothane, isoflurane, and enflurane have significant depressant effects on the spontaneous and epinephrine-altered automaticity of subsidiary atrial pacemakers. Depression of subsidiary atrial pacemaker automaticity was most apparent in dogs with muscarinic blockade.
Key words: Animal: anesthetized; conscious; dog. Anesthetics, volatile; enflurance; halothane; isoflurane. Heart: arrhythmias; autonomic regulation; sinoatrial node; subsidiary atrial pacemakers. Parasympathetic nervous system: acetylcholine. Sympathetic nervous system, catecholamines: epinephrine.
SUPPRESSION of normal SA node function with escape of subsidiary pacemakers may result in symptomatic bradycardia, and anesthetics may contribute to this phenomenon. Woehlck et al.  demonstrated in conscious and halothane-anesthetized dogs with intact sinoatrial (SA) nodes that increased vagal tone is the predominant mechanism for bradycardia and activation of latent atrial pacemakers during epinephrine-induced hypertension. In a subsequent study by Vicenzi et al.,  muscarinic blockade with methylatropine during enflurane anesthesia prevented epinephrine-induced and baroreflex-mediated bradycardia but not the concomitant pacemaker shifts. Intact muscarinic transmission was required for epinephrine-induced bradycardia and atrial pacemaker shifts during isoflurane anesthesia. .
Removal of the SA node is necessary to further characterize the properties of subsidiary pacemakers in conscious and anesthetized dogs, because it enables suppressed pacemakers to assume permanent control of the heart rhythm. It has been demonstrated that excision of the high rostral portion of the canine sulcus terminalis ultimately produced permanent atrial escape rhythms, after the subsidiary atrial pacemakers adapted to the lack of overdrive suppression by the SA node, [3–6] thus permitting studies in vivo over an extended period without competition from pacemakers with greater automatically. Although volatile anesthetic agents are known to modify SA nodal automatically through direct effects and the modification of autonomic reflexes, [7–10] their effects on spontaneous automaticity and recovery from overdrive suppression of subsidiary atrial pacemakers in vivo remain to be elucidated. Such knowledge should have relevance to the anesthetic management of patients with known or suspected SA node dysfunction.
This study attempts to show that subsidiary atrial pacemakers located at the midrostral portion of the sulcus terminalis permanently emerge after excision of the SA node. We will determine changes in the automaticity of these subsidiary pacemakers induced by the isolated and combined effects of epinephrine, the inhalation anesthetics halothane, isoflurane, and enflurane, as well as muscarinic blockade with methylatropine. Finally, we seek to demonstrate that, in comparison to previously published results from dogs with intact SA nodes, subsidiary atrial pacemakers have lower automaticity and that the anesthetics enhance these differences.
The methodology was similar to that reported previously, [1,2] with changes pertaining to the excision of the SA node. Briefly, with approval of the Institutional Animal Care Committee, a right thoracotomy was performed in eight mongrel dogs (18.5–21 kg). The method established by Randall et al. [4,11,12] was employed to excise atrial tissue (1.5 x 0.7 cm) from the rostral end of the sulcus terminalis and the junction with the superior vena cava (Figure 1). Successful excision of the SA node was verified intraoperatively by ECG recordings. Additionally, in some animals, histologic sections proved that the excised tissue contained the entire SA node. Bipolar electrode pairs were sutured to the epicardium of 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 along the sulcus terminalis to the junction with the inferior vena cava (Figure 1). A bipolar needle electrode was placed into the interventricular septum to record the His bundle electrogram.  Leads were tunneled subcutaneously and exited between the scapulae. A catheter for blood pressure measurement was placed into the abdominal aorta through a femoral artery, and the attached vascular access port was implanted subcutaneously. Frequent ECG recordings during the 3 weeks after SA node excision, but before testing with anesthetics, verified that dogs developed a stable subsidiary atrial pacemaker rhythm.
From the conscious, resting dog (control state), a 1-min paper record (4 ms + mm sup -1 resolution in data capture mode) was obtained, including surface ECG and electrograms from the His bundle, the patch electrodes, and the left atrial appendage. In addition, heart rate (HR) and blood pressure (SBP) were continuously measured. While heart rate provided the measure of spontaneous pacemaker automaticity, subsidiary atrial pacemaker recovery time (SAPRT) yielded a measure of automaticity in response to the stress of overdrive suppression. SAPRT was determined by pacing the right atrial appendage at constant rates of 200 and 250 beats/min for 30 s each.  Pulse duration was 2 ms and current was 20 mA. Measurements were performed on the His bundle tracing from the last pacing spike to the first spontaneous atrial signal after cessation of pacing. Conscious dogs were given infusions of 1 micro gram epinephrine *symbol* kg1*symbol* min1for 3 min through a forelimb vein. HR and SBP were recorded and allowed to return to baseline before continuing with the protocol (> 30 min). SAPRT was not measured during exposure to epinephrine because pacing interferes with measurements of spontaneous HR and the development of pacemaker shifts. In addition, epinephrine induces time-dependent and rapid alterations in HR, SBP, and autonomic tone, which may alter SAPRT during the 3-min infusion. After 30 min, the epinephrine infusion was repeated at a rate of 2 micro gram *symbol* kg1*symbol* min1. Subsequently, anesthesia was induced via mask with inhalation of either halothane, isoflurane, or enflurane in oxygen, and was followed by tracheal intubation. Each dog was anesthetized with all three agents, 5 days of rest were allowed between testing, and the order of the anesthetics was changed for each dog. End-tidal concentrations of the volatile agents and PCO2were monitored continuously by a mass spectrometer. Depth of anesthesia was maintained at 1.25 and finally 2 MAC.  Ventilation was controlled to keep end-tidal PCO2between 35 and 40 mmHg. Fluid maintenance was with Ringer's lactate solution (3–5 ml *symbol* kg1*symbol* h1). The body temperature was monitored and maintained at 37 degrees Celsius. Thirty minutes was allowed for equilibration at the desired MAC level before baseline recordings and data acquisition during epinephrine infusions were repeated. Experiments with complete peripheral muscarinic blockade were performed by intravenous injection of methylatropine nitrate 30 min before conscious state testing (AMN, 3 mg *symbol* kg1).
Normal distribution of all data was confirmed before further statistical analyses. SAPRT data were evaluated by univariate analysis of variance for repeated measures, followed by least-squares means tests. Analysis of covariance for repeated measures (univariate, no intercept model) was performed on HR data. Matching SBP measurements entered as the covariate. Individual comparisons of HR data were performed by least squares means tests with the covariate set to its mean. These comparisons represent (1) effects of epinephrine (as compared to no epinephrine), (2) effects of each anesthetic (compared to the corresponding conscious state condition), (3) effects of methylatropine (compared to the absence of methylatropine), and (4) comparisons between halothane, isoflurane, and enflurane. Significance was identified at P less or equal to 0.05. Data are shown as mean plus/minus SE.
Three weeks after removal of the SA node, the site of earliest atrial activation in conscious dogs without methylatropine or exposure to epinephrine (control state) had invariably shifted to the midrostal portion of the sulcus terminalis (third patch electrode, Figure 1). The surface ECG was of regular, supraventricular nature with positive P waves and showed minimal (< 25 msec) phasic variations with respiration. Detailed results of SAPRT, HR, and SBP are presented in Table 1and Table 2and Figure 2. Each dog served as its own control. The greatest correlation coefficient for any pair of HR data was 0.665, therefore correlation coefficients are not presented in detail. Individual comparisons were not performed on SBP data, because these data were entered as matching covariates in the analysis of HR.
Subsidiary Atrial Pacemaker Recovery Time
The first emerging, spontaneous atrial beat after cessation of pacing was always detected on the third patch electrode (midrostral sulcus terminalis). Halothane, isoflurane, and enflurane generally prolonged SAPRT, although not all anesthetics prolonged SAPRT under all conditions (Table 1). Without methylatropine, only SAPRT at 250 beats/min was prolonged, and during isoflurane, prolongation occurred only at 2 MAC. With methylatropine however, all three anesthetics, at 1.25 and 2 MAC and at 200 and 250 beats/min prolonged SAPRT. Isoflurane generally caused the least prolongation of SAPRT, and enflurane the greatest. Methylatropine had differential effects on SAPRT; it shortened SAPRT in the conscious state but prolonged it in the presence of each anesthetic, most notably with enflurane.
Spontaneous Pacemaker Automaticity
Effects of Epinephrine. Epinephrine affected spontaneous pacemaker automaticity in two ways, depending on the presence or absence of methylatropine (Figure 2and Table 2). Without methylatropine, exposure to epinephrine decreased the HR in the conscious state, with 1.25 MAC halothane, and with either concentration of isoflurane. However, epinephrine during enflurane anesthesia did not alter the HR. With methylatropine, epinephrine increased the HR under all test conditions.
Effects of Anesthetics. Without methylatropine, halothane did not alter the HR (Figure 2and Table 2). Isoflurane decreased the HR only when combined with epinephrine, and enflurane had variable effects on the HR. Without epinephrine, and at 1.25 MAC, enflurane decreased the HR more than isoflurane. During exposure to epinephrine, the decrease in HR was greater with isoflurane than with halothane or enflurane.
With methylatropine, all three anesthetics decreased the HR under any test condition. This decrease was greater with enflurane compared to halothane or isoflurane, when no epinephrine was administered. However, during exposure to epinephrine, the HR with halothane was greater than with isoflurane, which in turn was greater than the HR with enflurane.
Effects of Methylatropine. Methylatropine greatly increased the HR under all conscious state test conditions and increased the HR during halothane and isoflurane anesthesia when epinephrine was administered. Methylatropine, however, had no effect on the HR during 2 MAC enflurane and epinephrine. Without epinephrine, though, methylatropine decreased the HR during each anesthetic and at any MAC level tested.
Latent atrial pacemakers may substitute for the failing SA node in patients with sinus node dysfunction and sick sinus syndrome, preventing bradycardia and junctional escape rhythms.  Because these patients are prone to symptomatic bradycardia, escape rhythms and ectopic tachydysrhythmias, they frequently receive temporary or permanent pacemakers. Others may suffer symptoms of sinus node dysfunction only when subjected to anesthetics, catecholamines, anticholinergics, or autonomic changes. A recent study in 200 anesthetized adult patients revealed an approximate 40% incidence of sinus bradycardia, a 13% incidence of hemodynamically significant bradycardia, and 6% incidence of AV junctional rhythms.  The incidence of bradycardia in patients with manifest or latent sinus node dysfunction might be higher. Anticholinergic or sympathomimetic drugs administered for bradycardia might exacerbate subsidiary pacemaker instability, worsening dysrhythmias.  Therefore, we studied interactions of anesthetics, catecholamines, anticholinergics, and autonomic changes in subsidiary pacemakers with sinus node dysfunction. Briefly, we found that sinus node excision is followed by a stable subsidiary atrial pacemaker rhythm. Subsidiary pacemakers are slowed by vagal tone and have lower automaticity than the SA node with methylatropine but have equivalent rates with the addition of epinephrine. Anesthetics decrease automaticity in subsidiary atrial more than in SA nodal pacemakers, but some anesthetics have different effects on SA nodal and subsidiary pacemakers.
We previously reported that both direct and vagally mediated effects of anesthetics and catecholamines contribute to activation of subsidiary pacemakers in intact dogs. [1,2] During epinephrine-induced hypertension in the conscious state, and with halothane or isoflurane, the appearance of latent pacemakers accompanied a decrease in heart rate, presumably mediated by the baroreflex. [1,2] Epinephrine with enflurane appeared to enhance automaticity more in subsidiary atrial pacemakers than in the SA node, independent of the vagus.  Although subsidiary atrial pacemakers may be affected by the same processes causing sinus node dysfunction in patients with advanced sick sinus syndrome, we believe that SA node excision in dogs provides a reasonable model of less extensive disease of the atrial pacemaker complex.
In vivo electrophysiologic studies of subsidiary atrial pacemakers show wide variation in automaticity, [3–6,11,12,19–26] and different techniques of pacemaker destruction [11,21,22,27] or excision [3–6,11,12,19,20,23,24,26] may produce this variability. Pacemakers located at more rostral portions of the sulcus terminalis [3,11,24] closely resemble the SA node; those located at increasingly caudal portions of the sulcus terminalis [6,12,19,20,23,24,26] differ more from the SA node; and remote atrial pacemakers, not located on the sulcus terminalis, have much lower automaticity and may not be capable of assuming stable pacemaker function. [23,24] Adaptation processes following SA node excision reduce differences between SA nodal and subsidiary atrial pacemakers over time. [3–5,26] Because the above studies were conducted on both conscious animals [3–6,11,12,20–23,26] and animals under anesthesia, [19,24] only data from experiments with a similar location and extent of tissue excision, conducted under similar experimental conditions, should be compared to data from this study. [11,24].
Comparisons to Dogs with Intact SA Nodes
With the exception of SA node excision, the surgical procedure, instrumentation, recovery, testing protocol, and physical characteristics of the dogs were essentially the same in both groups. [1,2] When comparing the two groups of chronically instrumented dogs in their control states (the conscious, resting animal without exposure to epinephrine or methylatropine), heart rate, recovery times (SNRT and SAPRT respectively), and blood pressure did not differ between dogs with intact [1,2] and excised SA nodes. Under other conditions, particularly during anesthesia, substantial differences appeared. For the convenience of comparisons, summarized data from dogs with intact SA nodes [1,2] are presented.
Dogs with excised SA nodes often had a significantly higher blood pressure than dogs with intact SA nodes (Table 1). [1,2] Although similar in the control state, methylatropine caused the blood pressure to be greater in conscious dogs with excised SA nodes compared to dogs with intact SA nodes. [1,2] Further differences in blood pressure between groups occurred only during exposure to epinephrine. Because blood pressure and heart rate are variables that can influence each other via ventricular filling, stroke volume, and baro- and mechanoreceptor reflexes and because slopes and intercepts of regression lines between SBP data and HR data under various test conditions and for each pacemaker site are not known, this study takes this relation into account by including all SBP data from previously published experiments [1,2] as well as from the current ones as matching covariates in the analysis of heart rate measurements. Such analysis of covariance for repeated measures (univariate, no intercept model) removes the influence of matching blood pressure data from the comparisons to dogs with intact SA nodes statistically. In much the same way, this statistical procedure removes the effects of changes in blood pressure for within-group comparisons. .
Without methylatropine, no differences in heart rate were identified, but with methylatropine, the heart rate in conscious dogs without exposure to epinephrine was about 20 beats/min less (P < 0.05) than in dogs with intact SA nodes. [1,2] During 1.25 MAC isoflurane or enflurane, the heart rate during exposure to epinephrine tended to be less than the heart rate in intact dogs. [1,2] Blood pressure was usually greater in dogs with excised SA nodes, yet corresponding heart rates were similar, indicating that the particular value of heart rate is explained sufficiently by the effects of the matching covariate (blood pressure) and not by other possible effects. Without including the covariate, such distinction could not have been made. Furthermore, the regression slopes between the covariate and the variable were frequently inhomogeneous (unequal), indicating that subsidiary atrial pacemakers do not respond uniformly and parallel to changes in blood pressure.
Findings from this and prior studies [1,2] demonstrate that methylatropine exerted positive chronotropic properties on subsidiary atrial pacemakers in conscious dogs but negative chronotropic effects during inhalation anesthesia only in the absence of epinephrine. Although direct depressant effects of methylatropine compounds and facilitation of pacemaker shifts to sites with lower spontaneous automaticity have not yet been excluded, evidence from the literature, [29–31] combined with our own observations, [1,2] support the following hypothesis: Methylatropine is a strong peripheral muscarinic blocker, and at greater doses, such as used in this study, also a moderate ganglionic blocker. [29,31] The net effect on automaticity depends not only on the dose but also the degree and predominance of vagal or sympathetic tone. Conscious dogs have greater autonomic tone than anesthetized dogs, especially vagal. Therefore, methylatropine increases pacemaker automaticity in conscious dogs, because vagal pacemaker depression is completely blocked and the additional small reduction of sympathetic tone is insufficient for a negative chronotropic net effect. Because anesthetics generally reduce autonomic tone, the blockade of the remaining low degree of vagal tone and the additional reduction in sympathetic tone produces a net negative chronotropic effect. Methylatropine prolongs pacemaker recovery by the same mechanism. We concede that 3 mg *symbol* kg1methylatropine is intended to block the baro-reflex completely for a long duration in research animals; such doses are not used in clinical settings.
A study by Ardell et al.  in conscious dogs during treadmill exercise found significantly greater differences in automaticity between the SA node and subsidiary atrial pacemakers than our study. We attribute this difference to several factors. First, their more extensive tissue excisions may have removed more of the rostral pacemakers with greater automaticity. Second, their dogs had a wider range of weights (15–25 kg) compared to our study (18.5–21 kg), which could result in greater subject variability. Third, Ardell et al.'s study revealed two different sites of subsidiary atrial pacemakers controlling the heart rate. Finally, the statistical analyses were different, as Ardell et al.  neither reported detailed blood pressure measurements nor corrected by ANCOVA for possible effects on the heart rate.
In previous in vivo studies from our laboratory, resting conscious dogs with intact SA nodes exhibited considerable phasic variability of heart rate with respiration. [1,2] This respiratory sinus dysrhythmia was abolished by methylatropine and was drastically reduced during anesthesia with volatile agents. [1,2] In the current study, even without methylatropine, only minimal heart rate variations (< 25 ms) existed during the respiratory cycle of conscious dogs with excised SA nodes. The degree of vagal innervation correlates with vagal suppression of atrial pacemakers, and pacemaker suppression appears to decrease with increasing distance from the SA node along the sulcus terminalis. [1,2,6,32] This suggests that the respirophasic change of heart rate is much less in subsidiary atrial pacemakers because of their lesser vagal innervation. [11,42].
Similar pacemaker automaticity at comparable blood pressure in the control state of dogs with intact and excised SA nodes suggests that heart rate and pacemaker recovery times do not permit reliable conclusions on the nature and location of atrial pacemakers. Results from conscious dogs with methylatropine indicate a 10% reduction of the heart rate due to SA nodal excision but approximately 20–30% reduction during anesthesia. The low correlation coefficients between conscious and anesthetic test conditions for dogs with intact and excised SA nodes are further evidence for this effect.
Conscious animals with excised SA nodes developed epinephrine-induced hypertension and decreased heart rate, presumably mediated by the baroreflex. When methylatropine blocked muscarinic transmission to the heart, exposure to epinephrine increased heart rate. We conclude that subsidiary atrial pacemakers are suppressed by the vagus and that positive chronotropic effects of epinephrine are masked by intact vagal transmission. Although with methylatropine, the average heart rate of conscious intact dogs [1,2] was approximately 20 beats/min greater than that of dogs with excised SA nodes, epinephrine increased heart rate in both groups by roughly 30 beats/min and eliminated statistically significant differences between these two groups. This suggests that, in conscious dogs, atrial pacemakers at the midrostral sulcus terminalis have a similar capacity as the SA node to respond to exogenous epinephrine. [1,2] Rozanski et al. [3–5,26] found that, immediately after SA node removal in vitro, subsidiary atrial pacemaker activity depended entirely on exogenous norepinephrine stimulus. However, 4–11 months after SA node excision, subsidiary atrial pacemakers were less sensitive to norepinephrine than SA nodes in their study. [3–5,26] Our study was conducted 3–5 weeks after SA node excision, when hyperresponsiveness of subsidiary atrial pacemakers to catecholamines should exist, but in vivo, autonomic and humoral regulation of SA nodal and subsidiary atrial pacemakers probably account more for the observed differences in automaticity than intrinsic cellular properties between subsidiary atrial and SA nodal pacemakers.
Anesthetics may enhance differences in automaticity between the SA node and subsidiary atrial pacemakers,  and anesthetic-epinephrine interactions may increase the automaticity more in latent atrial pacemakers than the SA node. [2,33,34] Our results suggest that, with methylatropine, but without epinephrine, all three volatile agents reduce pacemaker automaticity in the midrostral sulcus terminalis more than in the SA node. [1,2] Enflurane depressed atrial automaticity more than the other agents.  With methylatropine, during isoflurane or enflurane anesthesia, epinephrine increased heart rate in dogs with excised SA nodes by 25–30% less than intact dogs.  We conclude that isoflurane and enflurane tend to depress the increase in rate secondary to epinephrine more in midrostral pacemakers after SA node excision than the SA rate.  Because no such differences occurred during halothane anesthesia in vivo, we suggest that halothane decreases the positive chronotropic response to epinephrine equally in subsidiary atrial pacemakers after SA node excision and in SA nodes.
Anesthetics generally prolonged SA node recovery times. [1,2] Because the spontaneous automaticity and pacemaker recovery times were reduced by anesthetics with or without methylatropine, but these anesthetics produced little or no change in automaticity without methylatropine, we conclude that the depressant effects of anesthetics can be partially masked by intact vagal transmission and that pacemaker recovery time appears to be a more sensitive parameter than the spontaneous pacemaker automaticity. It appears that anesthetics generally prolong recovery times more in subsidiary atrial pacemakers than in the SA node. [1,2].
Effects of the baroreflex are also present in anesthetized dogs without SA nodes. In this study, isoflurane appeared to preserve baroreflex-mediated suppression of subsidiary atrial pacemakers. Halothane at 1.25 MAC also preserves the baroreflex-induced decrease of spontaneous pacemaker automaticity, but it blunts this response at 2 MAC. Enflurane always abolished bradycardia secondary to epinephrine-induced hypertension. As noted before,  enflurane appears to attenuate the baroreflex more than the other volatile agents but is also the greatest direct depressant of automaticity.
In conclusion, after SA node excision, this study suggests the following. (1) Subsidiary atrial pacemakers located at the midrostral portion of the sulcus terminalis emerge as the primary cardiac pacemaker. (2) In conscious dogs without peripheral muscarinic blockade, subsidiary atrial pacemakers are indistinguishable from the SA node by spontaneous automaticity and recovery times. (3) Pacemakers at the midrostral sulcus terminalis are slowed by increased vagal tone. With methylatropine, their maximal spontaneous automaticity is less than that of the SA node. Exogenous epinephrine can eliminate this difference. (4) Halothane, isoflurane, and enflurane decrease spontaneous automaticity and recovery from overdrive suppression in subsidiary atrial more than in SA nodal pacemakers. (5) Isoflurane and enflurane decrease the responsiveness to epinephrine more in subsidiary than SA nodal pacemakers, whereas halothane affects these pacemakers equally. (6) Isoflurane preserves the baroreflex, which can result in profound bradycardia with epinephrine-induced hypertension. Halothane and enflurane attenuate baroreflex-mediated bradycardia more strongly. If these findings are applicable to humans, pacemaker depression by volatile anesthetics may produce profound bradycardia in patients with autonomic dysfunction and subsidiary pacemaker rhythms. Volatile anesthetics like isoflurane, which have little attenuation of the baroreceptor reflex, may contribute to severe bradycardia in the presence of hypertension.
The authors thank Dr. R. Hoffman, for his advice in statistical methods, and J. Krolikowski, for his technical assistance.