ORIGINAL ARTICLE
Autonomic Remodeling: How Atrial Fibrillation Begets Atrial Fibrillation in the First 24 Hours Ling Zhang, MD,* Sunny S. Po, MD, PhD,† Huan Wang, MD,* Benjamin J. Scherlag, PhD,† Hongliang Li, PhD,† Juan Sun, MD,* Yanmei Lu, MD,* Yitong Ma, PhD,* and Yuemei Hou, MD, PhD‡
Background: The mechanism(s) of how atrial fibrillation (AF)
sustains itself in the first 24 hours is not well understood.
Key Words: autonomic nervous system, atrial fibrillation, heart rate variability, remodeling (J Cardiovasc Pharmacol 2015;66:307–315)
Objective: We sought to investigate the role of autonomic remodeling in the first 24 hours of AF simulated by rapid atrial pacing (RAP).
Methods: Forty-eight rabbits were divided into 6 groups. One group (n = 8) was euthanized after baseline recordings. Another group (n = 8) did not receive RAP during the 24-hour period to serve as controls. In the other 4 groups, rabbits were euthanized after RAP for 4, 8, 12, or 24 hours (n = 8 for each). Before and after designated hours of RAP, atrial effective refractory period, heart rate variability, and left vagal and sympathetic nerve activity (VNA and SNA, respectively) were determined. The right and left atrial tissues were obtained for immunocytochemical analysis for growthassociated protein 43 (GAP43), tyrosine hydroxylase (TH), and choline acetyltransferase (ChAT). Results: RAP resulted in progressively shortened atrial effective refractory period and slower heart rate. In the first 12 hours of RAP, both SNA and VNA progressively increased. Then, VNA remained stably elevated but SNA began to attenuate. The high-frequency component and low-frequency/high-frequency ratio of heart rate variability followed the trend of VNA and SNA, respectively. The density of GAP43-positive, ChAT-positive, and TH-positive neural elements in the right and left atria was progressively higher with RAP. Conclusions: AF resulted in progressive autonomic remodeling, manifesting as nerve sprouting, sympathetic and vagal hyperinnervation. Autonomic remodeling may play an important role in sustaining AF in the first 24 hours. Received for publication January 20, 2015; accepted April 21, 2015. From the *Cardiovascular Center, First Affiliated Hospital of Xinjiang Medical University, Urumqi, China; †Heart Rhythm Institute, University of Oklahoma Health Sciences Center, Oklahoma City, OK; and ‡Department of Cardiovascular Diseases, The 6th People’s Hospital, Shanghai Jiaotong University, Shanghai, China. Supported in part by the International Science & Technology Cooperation Program of China (Grant No. 2011DFA32860) and fund of Xinjiang Key Laboratory for Medical Animal Model Research of China (Grant No. XJDX1103-2012-09). The authors report no conflicts of interest. S. S. Po and Y. Hou contributed equally to this work. Reprints: Yuemei Hou, MD, PhD, Department of Cardiovascular Diseases, The 6th People’s Hospital, Shanghai Jiaotong University, Shanghai 200025, China (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
INTRODUCTION
Atrial fibrillation (AF) is the most commonly encountered clinical arrhythmia. Since Wijffels et al proposed the concept of “AF begets AF,” it has evolved into a group of mechanisms working synergistically to maintain AF. Electrical remodeling is known to shorten the action potential duration by downregulating ionic currents such as the L-type Ca++ and Ito.1 Structural remodeling involves atrial fibrosis that causes anisotropic conduction and promotes reentry.2,3 Inflammation and disturbance of the redox potentials4 lead to changes of ionic current density and promote a substrate to maintain AF. However, most of the studies supporting the concept of AF begets AF were conducted on animals undergoing weeks of rapid pacing. How AF begets itself in the first 24 hours remains elusive. Previous studies demonstrated that hyperactivity of the cardiac autonomic nervous system (ANS) is critical in the initiation and maintenance of AF. In a canine model for paroxysmal AF simulated by rapid atrial pacing (RAP), the neural activity in the cardiac ganglionated plexi increased progressively as RAP continued.5,6 In other words, hyperactivity of the cardiac ANS and AF form a vicious cycle. The former can initiate AF, and the latter further enhances the activity of the cardiac ANS. In these studies, the neural activity was recorded only for 6 hours and sympathetic and vagal nerve activity (SNA and VNA, respectively) were not recorded separately. In this study, we sought to elucidate the mechanisms that allow AF to perpetuate itself in the first 24 hours. We recorded the atrial effective refractory period (ERP), heart rate variability (HRV), neural activity of the left SNA and VNA, and acquired immunohistochemical data at different time points (baseline, 4, 8, 12, and 24 hours of RAP).
METHODS Animal Preparation This study was approved by the Animal Use and Management Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University. Fifty-two male adult New Zealand white rabbits, weighing 2.0–2.5 kg, were
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
www.jcvp.org |
307
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Zhang et al
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
premedicated with ketamine/xylazine (35 mg/5 mg/kg) and restrained in the supine position. Anesthesia was also maintained by intraperitoneal injection of 20% urethane 5 mL as needed. The frequency of urethane administration and depth of anesthesia were adjusted by the appearance of muscle activity-related electrocardiographic (ECG) artifact. Standard surface ECG leads (I–aVF) were continuously monitored. Core body temperature was maintained at 38.58C 6 1.58C. The right jugular vein was dissected in the neck, and a 4 Fr multielectrode catheter was inserted into the high right atrium (HRA). The HRA site was used for pacing and determination of ERP. The left carotid artery was dissected and cannulated and connected to a pressure transducer to continuously monitor arterial blood pressure. The marginal ear vein was cannulated to allow fluid or medication administration (Fig. 1A).
Experiment Protocol Fifty-two rabbits were randomly assigned to 7 groups (Fig. 1B). Group 1 (n = 5): Left VNA and SNA were determined; then, after euthanasia, right (RA) and left atrial (LA) tissues were taken for immunohistochemistry (IHC), to serve as baseline data. Groups 2, 3, 4, and 5 (n = 8 for each group): Rabbits in these 4 groups were delivered RAP for 4, 6, 8, and 12 hours separately. At the end of 4, 6, 8, and 12 hours separately, left VNA and SNA were determined. Then, after euthanasia, RA and LA tissues were taken for IHC. Group 6 (n = 8): Rabbits were delivered RAP for 24 hours; at the baseline and at the end of 2, 4, 8, 12, and 24 hours, ERP, heart rate (HR), and indices of HRV were determined. At the end of 24 hours, left VNA and SNA were determined. Then, after euthanasia, RA and LA tissues were taken for IHC. Group 7 (n = 7): Sham RAP for 24 hours followed IHC. Rabbits did not receive RAP for 24 hours to serve as controls. All the indicators determined in Group 6 were similarly determined, to serve as controls.
RAP of the Right Atrium to Simulate AF Electrophysiological Studies In the baseline state, all rabbits were subjected to programmed electrical stimulation at the HRA site at 2 · diastolic threshold. Atrial pacing was delivered for eight beats (S1-S1 = 220 milliseconds), followed by a premature atrial stimulus. The S1–S2 interval was progressively decreased by 10 milliseconds and then 1 milliseconds until ERP was reached. RAP (pacing rate = 600 bpm) was stopped after 2, 4, 8, 12, or 24 hours for ERP measurement during sinus rhythm. AF was defined as an irregular atrial response faster than 500 bpm, lasting $5 seconds (Fig. 2A).
obtained at the baseline, 4, 8, 12, and 24 hours, respectively. HRV was measured by SD of normal-to-normal R-R interval, high-frequency (HF) component (0.15–0.40 Hz), lowfrequency (LF) component (0.04–0.15 Hz), and the LF/HF ratio.7,8
Sympathetic and Vagal Nerve Discharge Recording The cervical vagus nerve and the sympathetic nerve were identified and isolated using glass dissecting needles through a midline neck incision (Fig. 1B). They are all located posteromedial to carotid sheath and run between superior ganglia and middle ganglia, just beneath the cricoid cartilage about 0.5 cm. Nerves were covered by warm liquid paraffin cotton balls to maintain their moisture. The SNA and VNA were recorded using bipolar platinum recording electrodes through a BL-420 F Data Acquisition and Analysis System (ChengDu Technology & Market Co, Ltd, China). At each time point, after temporary cessation of RAP and restoration of sinus rhythm, SNA was recorded, followed by VNA (3 minutes for each). Data were recorded real time at a sampling rate of 5 kHz per channel and stored in a computer. High-pass (1 kHz) filtering of the time constant domain (0.001 seconds) and wavelet filtering (50 Hz) were used to eliminate the ECG signals before offline analysis. SNA, VNA, and SNA/VNA were analyzed for each time point (baseline, 4, 6, 8, 12, and 24 hours of RAP).9–11
Histological Studies After the predetermined duration of RAP for each group, rabbits were euthanized by injection of 50 mg/kg pentobarbital sodium, and the hearts were quickly removed. Myocardial samples from the LA and RA free wall at each time point were harvested and fixed in 10% neutral formalin for 48 hours and then embedded in paraffin blocks. Five-micrometer sections were cut from paraffin blocks of the RA and LA. The sections were stained with (1) growthassociated protein 43 (GAP43), a protein expressed in the growth cones of sprouting axons and a marker for nerve sprouting, (2) tyrosine hydroxylase (TH) to label sympathetic neural elements, and (3) choline acetyltransferase (ChAT) for cholinergic neural elements.12 Slides were examined with AxioVision software (AxioVision 4.1; Zeiss, Germany) using a CCD camera attached to an inverted light microscope with a 20· objective lens (Carl Zeiss, Germany). We determined nerve density by Image-Pro Plus 6.0 (Media Cybernetics). The nerve density was represented by the ratio of the positive nerves area to the total area (in mm2/mm2).
Statistical Analysis
In groups 6 and 7, after cessation of RAP for designated number of hours and after spontaneous termination of the induced AF, 5 minutes of Holter recording was obtained for HRV analysis by Century 3000 Holter Monitor (Biomedical Instruments, Co, Ltd). In the control group, the HRV was
Statistical analysis was performed using SPSS 17.0 (SPSS Inc, Chicago, IL). Data are expressed as mean 6 SD. Analysis of variance with post hoc Tukey’s test was used to compare continuous variables among different time points of RAP. Correlation analysis was undertaken using the Pearson’s test. Paired t-tests were used for pairwise comparisons of RA and LA. P values ,0.05 were considered statistically significant.
308
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
HRV Analysis
| www.jcvp.org
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Autonomic Nervous Remodeling in Paroxysmal AF
FIGURE 1. A, Flowchart of the experimental design. B, Schematic representation of animal preparation for recording of the BP, ERP, HRV, and cervical VNA and SNA.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.jcvp.org |
309
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Zhang et al
FIGURE 2. A, A typical example of AF after temporary discontinuation of 4 hours of RAP. The data are presented as mean 6 SD. In the RAP group, the atrial ERP shortened (B) and HR decreased (C) as the RAP time is increased. There are no significant changes in the control group. *P , 0.05 compared with the baseline.
RESULTS During RAP, the blood pressure was stable and no evident sign of heart failure was observed during the entire period of the experiments.
Effects of RAP on ERP and HR After RAP for 1 to 2 hours, AF induced. In the RAP pacing progress, average RR interval was determined by 20 ventricular beats at the baseline, 4, 8, 12, and 24 hours; the RR interval was progressively increased throughout the 24-hour pacing period (100 milliseconds, baseline; 119 6 35 milliseconds, 4 hours; 125 6 23 milliseconds, 8 hours; 156 6 42 milliseconds, 12 hours; 218 6 36 milliseconds,
24 hours). After RAP stop, AF cannot sustain for more than 5 seconds. As illustrated in Figure 2B, the ERP began to shorten significantly after 4 hours of RAP (93 6 14 milliseconds, baseline; 86 6 12 milliseconds, 4 hours; 81 6 11 milliseconds, 8 hours; 83 6 6 milliseconds, 12 hours; 75 6 10 milliseconds, 24 hours; P , 0.05 for all, compared with baseline). Similar to the trend of ERP change, the HR was progressively decreased throughout the 24-hour pacing period (240 6 10 bpm, baseline; 214 6 19 bpm, 8 hours; 184 6 11 bpm, 24 hours; P , 0.05, compared with baseline) (Fig. 2C). For control animals, ERP and HR did not change over the 24-hour period, indicating that the animals remained stable during the entire experiment.
FIGURE 3. Changes in HRV parameters during the 24 hours of RAP. The LF (A) and HF (B) components and the LF/HF ratio (C) showed an initial increase with RAP, followed by a decrease. In contrast, SD of normalto-normal R-R interval (D) showed a progressive decrease over the 24-hour period. *P , 0.05 compared with baseline.
310
| www.jcvp.org
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Autonomic Nervous Remodeling in Paroxysmal AF
FIGURE 4. Typical examples of neural recordings of the SNA and VNA up to 24 hours. Note that the scale is different in each panel to optimize visualization. Each panel shows the recordings of the SNA and VNA. Frequency of neural activity was calculated by manually counting the clusters of firing. For example, the window denoted by the 2 arrows in the panel of 6H was counted as 1 cluster. At 24 hours, there were 12 clusters of VNA and 5 clusters of SNA over 10 seconds, translating to 60 and 30 clusters per minute.
Effects of RAP on HRV Analysis Figure 3 illustrates the changes in HRV parameters determined by time-domain and frequency-domain analysis. There was a significant increase in the HF component from 4 hours to 12 hours; it then attenuated during the next 12 hours (P , 0.05, compared with baseline; Fig. 3A). As the RAP continued, the LF component was significantly increased from 25 6 3 ms2 in the baseline state to 47 6 5 ms2 and 65 6 5 ms2 at the end of 4 hours and 8 hours, respectively (P , 0.05, Fig. 3B). Then, Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
LF decreased to 55 6 5 ms2 at the end of 12 hours and to 36 6 3 ms2 at the end of pacing. With RAP, LF/HF ratio increased during the first 12 hours (2.72 6 0.22, baseline; 4.23 6 0.37, 4 hours; 4.21 6 0.15, 8 hours; 3.38 6 0.28, 12 hours; P , 0.05 for all, compared with baseline; Fig. 3C). The LF/HF ratio reverted back to the baseline at 24 hours. SD of normal-to-normal R-R interval shortened progressively as the RAP duration continued to diminish (7.50 6 0.41 milliseconds at baseline; 6.30 6 0.53 milliseconds at www.jcvp.org |
311
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Zhang et al
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
FIGURE 5. Changes in SNA and VNA during the 24 hours of RAP. Frequency (A) and amplitude (B) of the SNA and VNA changes. C, The ratio of SNA/VNA amplitude over a 24-hour period. D, Correlation between the SNA and VNA. The clusters per minute (CPM) of VNA showed an initial increase with RAP, followed by a decrease. In contrast, the CPM of SNA gradually decreases and reaches a stable level at 6 hours. VNA has a higher CPM than SNA over the 24-hour period (A). The amplitude of VNA and SNA discharges progressively increased with RAP in the first 8 hours, and VNA remained stably elevated afterward; in contrast, SNA decreased at 24 hours. Of note, the SNA and VNA in the control group did not change over 24 hours. C, Increased VNA was associated with increased SNA from 2 to 12 hours. A good correlation between SNA and VNA was also found (r = 0.816 6 0.105, P , 0.001; D). *P , 0.05 compared with the baseline; #P , 0.05, compared with baseline in RAP group-VNA; ##P , 0.01, comparison between VNA and SNA.
4 hours; 5.65 6 0.54 milliseconds at 8 hours; 5.33 6 1.00 milliseconds at 12 hours; 3.77 6 0.43 milliseconds at 24 hours, P , 0.05, Fig. 3D). No time-dependent change was noted in the control group.
Effects of RAP on SNA and VNA Recording of the SNA and VNA was successful in all rabbits. In the baseline state, the VNA and SNA showed low-amplitude activities and increased progressively with RAP (Fig. 4). The VNA frequency and amplitude (Fig. 5A–B) showed an initial increase with RAP and plateaued between 12 and 24 hours. In contrast to VNA, the frequency of SNA did not change over time (Fig. 5A), but the amplitude of SNA increased from 9.89 6 2.48 mV to 21.41 6 3.84 mV at 4 hours and 89.36 6 8.98 at 8 hours and then remained stable from 8 to 12 hours. It was then followed by SNA withdrawal to 70.6 6 4.23 mV during 12–24 hours (Fig. 5B). The amplitude of VNA discharges progressively increased with RAP in the first 8 hours and remained stably elevated afterward. Of note, the SNA and VNA in the control group did not change over 24 hours. Figure 5C showed that increased VNA was associated with increased SNA from 2 to 12 hours. A good correlation between SNA and VNA was also found (r = 0.816 6 0.105, P , 0.001; Fig. 5D).
312
| www.jcvp.org
Histological Studies Immunohistochemical staining showed that GAP43-, ChAT-, and TH-immunoreactive neural elements were present in RA and LA tissues at 4, 8, 12, and 24 hours (Fig. 6A). In the baseline state, the RA had a slightly higher GAP43-positive and TH-positive nerve density than the LA, while the distribution of ChAT-positive nerves was reversed. The LA also contained more parasympathetic nerves than the RA. As RAP continued, GAP43-, ChAT-, and TH-positive nerve density progressively increased (P , 0.05; compared with control, Fig. 6B–D). Differential expression of TH-positive and ChAT-positive nerves between the RA and LA was also observed.
DISCUSSION Main Findings
To the best of our knowledge, this study is the first one to assess the interactions between electrical and autonomic remodeling during the first 24 hours after AF initiation. We provided experimental evidence for the presence of both electrical and autonomic remodeling. The former is characterized by progressive shortening of the ERP along with progressive slowing of the sinus rate (inverse of rate adaptation). The latter was characterized by increased VNA and SNA, corresponding changes in HRV parameters, and Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Autonomic Nervous Remodeling in Paroxysmal AF
FIGURE 6. A, In the baseline state, the RA had a slightly higher GAP43-positive and TH-positive nerve density than the LA, while the distribution of ChAT-positive nerves was reversed. The LA also contained more parasympathetic nerves than the RA. As RAP continued, GAP43-, ChAT-, and TH-positive nerve density progressively increased compared with control. Differential expression of TH-positive and ChATpositive nerves between the RA and LA was also observed. Immunohistochemical staining for GAP43, ChAT, and TH antibodies in atrial tissue (dark brown stain) in control rabbits and in rabbits undergoing 4, 8, 12, and 24 hours of RAP. As RAP continued, GAP43 (B), TH(+) (C), and ChAT(+) (D) nerve density progressively increased. *P , 0.05 compared with the BS in the RA; #P , 0.05 compared with the BS in the LA; §P , 0.001, LA compared with the RA in the same group. Magnification, ·200.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.jcvp.org |
313
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Zhang et al
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
increased sympathetic and vagal innervation of the atria. Notably, ERP began to shorten and innervation began to increase after only 4 hours of RAP, indicating that both electrical and autonomic remodeling are necessary in maintaining AF in the first few hours after its initiation.
Electrical Remodeling to Perpetuate AF In 1995, 2 independent experimental studies reported that tachycardia-induced atrial electrical remodeling created a substrate for persistent AF, leading to the concept of AF begets AF.13,14 Both acute (hours) and chronic (weeks to months) RAP induced shortening of the atrial ERP and loss of rate adaptation.15–17 In this study, we found that with RAP, more pronounced ERP shortening was accompanied by slower HR, indicating that the physiological rate adaptation of the refractory period was inverted during the first 24 hours of simulated AF. Rapid atrial rate such as AF induced intracellular Ca2+ overload, followed by inactivation of L-type Ca++ channels. The reduced L-type Ca++ current shortens both the action potential duration and ERP.18 The transient outward potassium current (Ito) is reduced along with increased inward rectifier potassium current (IKI) and the acetylcholineactivated potassium current (IKACh). All of these factors contribute to the decreased action potential duration and rate adaptation changes.1 However, complete reversal of electrical remodeling can occur within a few days after restoration of sinus rhythm despite a prolonged period of AF (months).14,19,20 Without structural remodeling, which typically takes place after several weeks of persistent AF or RAP,18 electrical remodeling alone may not be able to maintain AF in the first 24 hours after AF was initiated.
Autonomic Remodeling to Perpetuate AF
Clinical and experimental studies have indicated fluctuation of the sympathetic and vagal tone preceding the onset of RAP.21,22 Vagal hyperactivity alone can shorten the ERP and increase ERP dispersion,23–25 promoting a substrate for reentry. Zhou et al proposed an “octopus hypothesis,” which states that the cardiac ANS is like an octopus. The head and tentacles of the octopus are the ganglionated plexi and autonomic nerves, respectively.26 This study is the first one to demonstrate specific patterns of autonomic nerve activity within the first 24 hours of AF. We found a progressive increase in the HF component and LF/HF ratio of HRV and an increase in SNA and VNA as well as the density of both the sympathetic and vagal nerves. These progressive changes were observed as early as 4 hours after RAP and continued until 12 hours. Between 12 and 24 hours, the nerve density continued to increase but the SNA and VNA decreased. At 24 hours, it reached a vagalpredominant state, evidence by a slower sinus rate (Fig. 2D) and a low SNA/VNA ratio (Fig. 5). We do not have experimental evidence to explain such dissociations between the continuously increased nerve density and attenuated nerve activity. We hypothesized that autonomic hyperinnervation may triggered an autonomic reflex to reduce the SNA and VNA as a protective mechanism. However, hyperinnervation and sprouting of the atrial autonomic nerves have set the stage for perpetuating AF beyond 24 hours. For example, the
314
| www.jcvp.org
hyperactive vagal activity at 24 hours can promote reentry by abbreviating the ERP and enhancing ERP dispersion. Stretch of the pulmonary veins or atrium can further abbreviate the ERP. Additional sympathetic stimulation by higher circulating catecholamine levels can easily facilitate the reinitiation and maintenance of AF beyond 24 hours. Recent experimental studies demonstrated that atrial nerve sprouting and sympathetic hyperinnervation provide a substrate for the development of AF.27–29 Atrial cholinergic innervation and peptidergic innervation also increased in animals undergoing long-term RAP and play an important role in the arrhythmogenesis of AF.29,30 The relative abundance of sympathetic hyperinnervation between the RA and LA often depends on the experimental model in which the electrical current delivered by the pacing lead is known to accelerate axonal regeneration31–33 and may result in nerve sprouting. In this study, sympathetic nerve sprouting appeared to be more pronounced in the RA than in the LA (Fig. 6). It is possible that this observation might be caused by rapid RA pacing. Regardless of the relative abundance of nerve sprouting between the RA and LA, heterogeneous atrial autonomic hyperinnervation is a common finding in patients with persistent or chronic AF.34–36 This study provides additional evidence to indicate the importance of the role of sympathetic and cholinergic hyperinnervation in the maintenance of AF.
Clinical Perspectives In this study, we focused on the mechanisms that maintain AF in the first 24 hours. Using a 24-hour RAP model in rabbits to simulate AF, we verified the presence of electrical remodeling as early as 4 hours. The SNA and VNA as well as the nerve density of the sympathetic and vagal nerves significantly began to progressively increase as early as 4 hours. So, we assume that autonomic remodeling, working synergistically with electrical remodeling, is important in maintaining AF and perpetuating AF beyond 24 hours. As the population ages, the incidence of AF is projected to sharply increase in the next 2–3 decades. As AF is a progressive disease, early interventions may prevent the development of irreversible changes such as fibrosis. This study calls for therapies designed specifically to inhibit autonomic remodeling in order to break the vicious cycle formed by electrical and autonomic remodeling. Early termination of AF will have a substantially positive impact on reducing the morbidities and mortality caused by more persistent forms of AF.
Study Limitations The stimulus strength of this study was relatively higher (twice diastolic threshold). In fact, the mean threshold is 0.35 6 0.10 mV. The higher stimulus strength might decrease the ERP in some certain. Because of the limitation of the neural recording device, VNA and SNA were not recorded continuously; at each time point, the SNA was recorded 3 minutes earlier than the VNA. Therefore, the exact temporal association between SNA and VNA was not studied. However, our experimental findings clearly illustrate the time course of the changes in SNA and VNA as well as the importance of progressive autonomic remodeling in AF maintenance. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
This study was conducted under general anesthesia that may introduce a confounding factor into the measurement of the ANS function, leading to the observation that the HRV data did not correlate well with the SNA and VNA after 12 hours. Also, ketamine might cause arrhythmias. However, all animals were under the same type of anesthesia, and the HR and HRV parameters were stable over 24 hours in the control group (Figs. 2–3), indicating that the changes in the ANS during the 24 hours of RAP are not merely artifacts from changes of ANS activity during anesthesia. To observe arrhythmia caused by ketamine, we have sham RAP group in which rabbits did not receive RAP for 24 hours to serve as controls; in this group, we have not observed any arrhythmia.
CONCLUSIONS Progressive autonomic remodeling manifested as autonomic nerve sprouting, sympathetic and vagal hyperinnervation, and a shift of the sympathetic/vagal balance in the first 24 hours of AF simulated by RAP. Working synergistically with electrical remodeling, autonomic remodeling may be a critical element in maintaining AF in the first 24 hours after its initiation. ACKNOWLEDGMENTS The authors are grateful to the Xinjiang Animal Research Center Facility staff Mei Ma and Mingjun Duan, Qing Wei, and Chun Zhang for their assistance. REFERENCES 1. Bosch RF, Zeng X, Grammer JB, et al. Ionic mechanisms of electrical remodeling in human atrial fibrillation. Cardiovasc Res. 1999;44: 121–131. 2. Verheule S, Wilson E, Everett T, et al. Alterations in atrial electrophysiology and tissue structure in a canine model of chronic atrial dilatation due to mitral regurgitation. Circulation. 2003;107:2615–2622. 3. Verheule S, Wilson E, Banthia S, et al. Direction-dependent conduction abnormalities in a canine model of atrial fibrillation due to chronic atrial dilatation. Am J Physiol Heart Circ Physiol. 2004;287:H634–H644. 4. Anderson EJ, Efird JT, Davies SW, et al. Monoamine oxidase is a major determinant of redox balance in human atrial myocardium and is associated with postoperative atrial fibrillation. J Am Heart Assoc. 2014;3:e000713. 5. Yu L, Scherlag BJ, Li S, et al. Low-level transcutaneous electrical stimulation of the auricular branch of the vagus nerve: a noninvasive approach to treat the initial phase of atrial fibrillation. Heart Rhythm. 2013;10:428–435. 6. Yu L, Scherlag BJ, Sha Y, et al. Interactions between atrial electrical remodeling and autonomic remodeling: how to break the vicious cycle. Heart Rhythm. 2012;9:804–809. 7. Xhyheri B, Manfrini O, Mazzolini M, et al. Heart rate variability today. Prog Cardiovasc Dis. 2012;55:321–331. 8. Camm AJ, Malik M, Bigger JT. Heart rate variability—standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation. 1996;93:1043–1065. 9. Jung BC, Dave AS, Tan AY, et al. Circadian variations of stellate ganglion nerve activity in ambulatory dogs. Heart Rhythm. 2006;3:78–85. 10. Piccirillo G, Magri D, Ogawa M, et al. Autonomic nervous system activity measured directly and QT interval variability in normal and pacing-induced tachycardia heart failure dogs. J Am Coll Cardiol. 2009;54:840–850. 11. Shen MJ, Choi EK, Tan AY, et al. Patterns of baseline autonomic nerve activity and the development of pacing-induced sustained atrial fibrillation. Heart Rhythm. 2011;8:583–589.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Autonomic Nervous Remodeling in Paroxysmal AF
12. Tan AY, Zhou S, Ogawa M, et al. Neural mechanisms of paroxysmal atrial fibrillation and paroxysmal atrial tachycardia in ambulatory canines. Circulation. 2008;118:916–925. 13. Morillo CA, Klein GJ, Jones DL, et al. Chronic rapid atrial pacing. Structural, functional, and electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation. 1995;91:1588–1595. 14. Wijffels MC, Kirchhof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation. 1995;92:1954–1968. 15. Lu Z, Scherlag BJ, Lin J, et al. Atrial fibrillation begets atrial fibrillation: autonomic mechanism for atrial electrical remodeling induced by shortterm rapid atrial pacing. Circ Arrhythm Electrophysiol. 2008;1:184–192. 16. Attuel P, Childers R, Cauchemez B, et al. Failure in the rate adaptation of the atrial refractory period: its relationship to vulnerability. Int J Cardiol. 1982;2:179–197. 17. Boutjdir M, Le Heuzey JY, Lavergne T, et al. Inhomogeneity of cellular refractoriness in human atrium: factor of arrhythmia? Pacing Clin Electrophysiol. 1986;9:1095–1100. 18. Gaspo R, Bosch RF, Talajic M, et al. Functional mechanisms underlying tachycardia-induced sustained atrial fibrillation in a chronic dog model. Circulation. 1997;96:4027–4035. 19. Resnekov L, McDonald L. Appraisal of electroconversion in treatment of cardiac dysrhythmias. Br Heart J. 1968;30:786–811. 20. Bjerkelund C, Orning OM. An evaluation of DC shock treatment of atrial arrhythmias. Acta Med Scand. 1968;184:481–491. 21. Bettoni M, Zimmermann M. Autonomic tone variations before the onset of paroxysmal atrial fibrillation. Circulation. 2002;105:2753–2759. 22. Fioranelli M, Piccoli M, Mileto GM, et al. Analysis of heart rate variability five minutes before the onset of paroxysmal atrial fibrillation. Pacing Clin Electrophysiol. 1999;22:743–749. 23. Goldberger AL, Pavelec RS. Vagally-mediated atrial fibrillation in dogs: conversion with bretylium tosylate. Int J Cardiol. 1986;13:47–55. 24. Oliveira M, da Silva MN, Geraldes V, et al. Acute vagal modulation of electrophysiology of the atrial and pulmonary veins increases vulnerability to atrial fibrillation. Exp Physiol. 2011;96:125–133. 25. Oliveira M, Silva MN, Geraldes V, et al. Effects of vagal stimulation on induction and termination of atrial fibrillation in an in vivo rabbit heart model. Rev Port Cardiol. 2010;29:375–389. 26. Zhou JSB, Edwards J, Jackman WM, et al. Gradients of atrial refractoriness and inducibility of atrial fibrillation due to stimulation of ganglionated plexi. J Cardiovasc Electrophysiol. 2007;18:83–90. 27. Jayachandran JV, Sih HJ, Winkle W, et al. Atrial fibrillation produced by prolonged rapid atrial pacing is associated with heterogeneous changes in atrial sympathetic innervation. Circulation. 2000;101:1185–1191. 28. Chang CM, Wu TJ, Zhou S, et al. Nerve sprouting and sympathetic hyperinnervation in a canine model of atrial fibrillation produced by prolonged right atrial pacing. Circulation. 2001;103:22–25. 29. Yu Y, Liu L, Jiang JY, et al. Parasympathetic and substance P-immunoreactive nerve denervation in atrial fibrillation models. Cardiovasc Pathol. 2012;21:39–45. 30. Qu X, Yu Y, Jiang J, et al. Variance of peptidic nerve innervation in a canine model of atrial fibrillation produced by prolonged atrial pacing. Pacing Clin Electrophysiol. 2008;31:207–213. 31. Nix WA, Hopf HC. Electrical stimulation of regenerating nerve and its effect on motor recovery. Brain Res. 1983;272:21–25. 32. Pockett S, Gavin RM. Acceleration of peripheral nerve regeneration after crush injury in rat. Neurosci Lett. 1985;59:221–224. 33. Manivannan S, Terakawa S. Rapid sprouting of filopodia in nerve terminals of chromaffin cells, PC12 cells, and dorsal root neurons induced by electrical stimulation. J Neurosci. 1994;14:5917–5928. 34. Gould PA, Yii M, McLean C, et al. Evidence for increased atrial sympathetic innervation in persistent human atrial fibrillation. Pacing Clin Electrophysiol. 2006;29:821–829. 35. Tan AY, Li HM, Wachsmann-Hogiu S, et al. Autonomic innervation and segmental muscular disconnections at the human pulmonary vein-atrial junction—implications for catheter ablation of atrial-pulmonary vein junction. J Am Coll Cardiol. 2006;48:132–143. 36. Li Y, Lu Z, Tang Q, et al. The increase in sympathetic nerve density in the atrium facilitates atrial fibrillation in patients with rheumatic heart disease. Int J Cardiol. 2013;165:174–178.
www.jcvp.org |
315
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Autonomic Remodeling: How Atrial Fibrillation Begets Atrial Fibrillation in the First 24 Hours Ling Zhang, MD,* Sunny S. Po, MD, PhD,† Huan Wang, MD,* Benjamin J. Scherlag, PhD,† Hongliang Li, PhD,† Juan Sun, MD,* Yanmei Lu, MD,* Yitong Ma, PhD,* and Yuemei Hou, MD, PhD‡
Background: The mechanism(s) of how atrial fibrillation (AF)
sustains itself in the first 24 hours is not well understood.
Key Words: autonomic nervous system, atrial fibrillation, heart rate variability, remodeling (J Cardiovasc Pharmacol 2015;66:307–315)
Objective: We sought to investigate the role of autonomic remodeling in the first 24 hours of AF simulated by rapid atrial pacing (RAP).
Methods: Forty-eight rabbits were divided into 6 groups. One group (n = 8) was euthanized after baseline recordings. Another group (n = 8) did not receive RAP during the 24-hour period to serve as controls. In the other 4 groups, rabbits were euthanized after RAP for 4, 8, 12, or 24 hours (n = 8 for each). Before and after designated hours of RAP, atrial effective refractory period, heart rate variability, and left vagal and sympathetic nerve activity (VNA and SNA, respectively) were determined. The right and left atrial tissues were obtained for immunocytochemical analysis for growthassociated protein 43 (GAP43), tyrosine hydroxylase (TH), and choline acetyltransferase (ChAT). Results: RAP resulted in progressively shortened atrial effective refractory period and slower heart rate. In the first 12 hours of RAP, both SNA and VNA progressively increased. Then, VNA remained stably elevated but SNA began to attenuate. The high-frequency component and low-frequency/high-frequency ratio of heart rate variability followed the trend of VNA and SNA, respectively. The density of GAP43-positive, ChAT-positive, and TH-positive neural elements in the right and left atria was progressively higher with RAP. Conclusions: AF resulted in progressive autonomic remodeling, manifesting as nerve sprouting, sympathetic and vagal hyperinnervation. Autonomic remodeling may play an important role in sustaining AF in the first 24 hours. Received for publication January 20, 2015; accepted April 21, 2015. From the *Cardiovascular Center, First Affiliated Hospital of Xinjiang Medical University, Urumqi, China; †Heart Rhythm Institute, University of Oklahoma Health Sciences Center, Oklahoma City, OK; and ‡Department of Cardiovascular Diseases, The 6th People’s Hospital, Shanghai Jiaotong University, Shanghai, China. Supported in part by the International Science & Technology Cooperation Program of China (Grant No. 2011DFA32860) and fund of Xinjiang Key Laboratory for Medical Animal Model Research of China (Grant No. XJDX1103-2012-09). The authors report no conflicts of interest. S. S. Po and Y. Hou contributed equally to this work. Reprints: Yuemei Hou, MD, PhD, Department of Cardiovascular Diseases, The 6th People’s Hospital, Shanghai Jiaotong University, Shanghai 200025, China (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
INTRODUCTION
Atrial fibrillation (AF) is the most commonly encountered clinical arrhythmia. Since Wijffels et al proposed the concept of “AF begets AF,” it has evolved into a group of mechanisms working synergistically to maintain AF. Electrical remodeling is known to shorten the action potential duration by downregulating ionic currents such as the L-type Ca++ and Ito.1 Structural remodeling involves atrial fibrosis that causes anisotropic conduction and promotes reentry.2,3 Inflammation and disturbance of the redox potentials4 lead to changes of ionic current density and promote a substrate to maintain AF. However, most of the studies supporting the concept of AF begets AF were conducted on animals undergoing weeks of rapid pacing. How AF begets itself in the first 24 hours remains elusive. Previous studies demonstrated that hyperactivity of the cardiac autonomic nervous system (ANS) is critical in the initiation and maintenance of AF. In a canine model for paroxysmal AF simulated by rapid atrial pacing (RAP), the neural activity in the cardiac ganglionated plexi increased progressively as RAP continued.5,6 In other words, hyperactivity of the cardiac ANS and AF form a vicious cycle. The former can initiate AF, and the latter further enhances the activity of the cardiac ANS. In these studies, the neural activity was recorded only for 6 hours and sympathetic and vagal nerve activity (SNA and VNA, respectively) were not recorded separately. In this study, we sought to elucidate the mechanisms that allow AF to perpetuate itself in the first 24 hours. We recorded the atrial effective refractory period (ERP), heart rate variability (HRV), neural activity of the left SNA and VNA, and acquired immunohistochemical data at different time points (baseline, 4, 8, 12, and 24 hours of RAP).
METHODS Animal Preparation This study was approved by the Animal Use and Management Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University. Fifty-two male adult New Zealand white rabbits, weighing 2.0–2.5 kg, were
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
www.jcvp.org |
307
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Zhang et al
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
premedicated with ketamine/xylazine (35 mg/5 mg/kg) and restrained in the supine position. Anesthesia was also maintained by intraperitoneal injection of 20% urethane 5 mL as needed. The frequency of urethane administration and depth of anesthesia were adjusted by the appearance of muscle activity-related electrocardiographic (ECG) artifact. Standard surface ECG leads (I–aVF) were continuously monitored. Core body temperature was maintained at 38.58C 6 1.58C. The right jugular vein was dissected in the neck, and a 4 Fr multielectrode catheter was inserted into the high right atrium (HRA). The HRA site was used for pacing and determination of ERP. The left carotid artery was dissected and cannulated and connected to a pressure transducer to continuously monitor arterial blood pressure. The marginal ear vein was cannulated to allow fluid or medication administration (Fig. 1A).
Experiment Protocol Fifty-two rabbits were randomly assigned to 7 groups (Fig. 1B). Group 1 (n = 5): Left VNA and SNA were determined; then, after euthanasia, right (RA) and left atrial (LA) tissues were taken for immunohistochemistry (IHC), to serve as baseline data. Groups 2, 3, 4, and 5 (n = 8 for each group): Rabbits in these 4 groups were delivered RAP for 4, 6, 8, and 12 hours separately. At the end of 4, 6, 8, and 12 hours separately, left VNA and SNA were determined. Then, after euthanasia, RA and LA tissues were taken for IHC. Group 6 (n = 8): Rabbits were delivered RAP for 24 hours; at the baseline and at the end of 2, 4, 8, 12, and 24 hours, ERP, heart rate (HR), and indices of HRV were determined. At the end of 24 hours, left VNA and SNA were determined. Then, after euthanasia, RA and LA tissues were taken for IHC. Group 7 (n = 7): Sham RAP for 24 hours followed IHC. Rabbits did not receive RAP for 24 hours to serve as controls. All the indicators determined in Group 6 were similarly determined, to serve as controls.
RAP of the Right Atrium to Simulate AF Electrophysiological Studies In the baseline state, all rabbits were subjected to programmed electrical stimulation at the HRA site at 2 · diastolic threshold. Atrial pacing was delivered for eight beats (S1-S1 = 220 milliseconds), followed by a premature atrial stimulus. The S1–S2 interval was progressively decreased by 10 milliseconds and then 1 milliseconds until ERP was reached. RAP (pacing rate = 600 bpm) was stopped after 2, 4, 8, 12, or 24 hours for ERP measurement during sinus rhythm. AF was defined as an irregular atrial response faster than 500 bpm, lasting $5 seconds (Fig. 2A).
obtained at the baseline, 4, 8, 12, and 24 hours, respectively. HRV was measured by SD of normal-to-normal R-R interval, high-frequency (HF) component (0.15–0.40 Hz), lowfrequency (LF) component (0.04–0.15 Hz), and the LF/HF ratio.7,8
Sympathetic and Vagal Nerve Discharge Recording The cervical vagus nerve and the sympathetic nerve were identified and isolated using glass dissecting needles through a midline neck incision (Fig. 1B). They are all located posteromedial to carotid sheath and run between superior ganglia and middle ganglia, just beneath the cricoid cartilage about 0.5 cm. Nerves were covered by warm liquid paraffin cotton balls to maintain their moisture. The SNA and VNA were recorded using bipolar platinum recording electrodes through a BL-420 F Data Acquisition and Analysis System (ChengDu Technology & Market Co, Ltd, China). At each time point, after temporary cessation of RAP and restoration of sinus rhythm, SNA was recorded, followed by VNA (3 minutes for each). Data were recorded real time at a sampling rate of 5 kHz per channel and stored in a computer. High-pass (1 kHz) filtering of the time constant domain (0.001 seconds) and wavelet filtering (50 Hz) were used to eliminate the ECG signals before offline analysis. SNA, VNA, and SNA/VNA were analyzed for each time point (baseline, 4, 6, 8, 12, and 24 hours of RAP).9–11
Histological Studies After the predetermined duration of RAP for each group, rabbits were euthanized by injection of 50 mg/kg pentobarbital sodium, and the hearts were quickly removed. Myocardial samples from the LA and RA free wall at each time point were harvested and fixed in 10% neutral formalin for 48 hours and then embedded in paraffin blocks. Five-micrometer sections were cut from paraffin blocks of the RA and LA. The sections were stained with (1) growthassociated protein 43 (GAP43), a protein expressed in the growth cones of sprouting axons and a marker for nerve sprouting, (2) tyrosine hydroxylase (TH) to label sympathetic neural elements, and (3) choline acetyltransferase (ChAT) for cholinergic neural elements.12 Slides were examined with AxioVision software (AxioVision 4.1; Zeiss, Germany) using a CCD camera attached to an inverted light microscope with a 20· objective lens (Carl Zeiss, Germany). We determined nerve density by Image-Pro Plus 6.0 (Media Cybernetics). The nerve density was represented by the ratio of the positive nerves area to the total area (in mm2/mm2).
Statistical Analysis
In groups 6 and 7, after cessation of RAP for designated number of hours and after spontaneous termination of the induced AF, 5 minutes of Holter recording was obtained for HRV analysis by Century 3000 Holter Monitor (Biomedical Instruments, Co, Ltd). In the control group, the HRV was
Statistical analysis was performed using SPSS 17.0 (SPSS Inc, Chicago, IL). Data are expressed as mean 6 SD. Analysis of variance with post hoc Tukey’s test was used to compare continuous variables among different time points of RAP. Correlation analysis was undertaken using the Pearson’s test. Paired t-tests were used for pairwise comparisons of RA and LA. P values ,0.05 were considered statistically significant.
308
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
HRV Analysis
| www.jcvp.org
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Autonomic Nervous Remodeling in Paroxysmal AF
FIGURE 1. A, Flowchart of the experimental design. B, Schematic representation of animal preparation for recording of the BP, ERP, HRV, and cervical VNA and SNA.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.jcvp.org |
309
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Zhang et al
FIGURE 2. A, A typical example of AF after temporary discontinuation of 4 hours of RAP. The data are presented as mean 6 SD. In the RAP group, the atrial ERP shortened (B) and HR decreased (C) as the RAP time is increased. There are no significant changes in the control group. *P , 0.05 compared with the baseline.
RESULTS During RAP, the blood pressure was stable and no evident sign of heart failure was observed during the entire period of the experiments.
Effects of RAP on ERP and HR After RAP for 1 to 2 hours, AF induced. In the RAP pacing progress, average RR interval was determined by 20 ventricular beats at the baseline, 4, 8, 12, and 24 hours; the RR interval was progressively increased throughout the 24-hour pacing period (100 milliseconds, baseline; 119 6 35 milliseconds, 4 hours; 125 6 23 milliseconds, 8 hours; 156 6 42 milliseconds, 12 hours; 218 6 36 milliseconds,
24 hours). After RAP stop, AF cannot sustain for more than 5 seconds. As illustrated in Figure 2B, the ERP began to shorten significantly after 4 hours of RAP (93 6 14 milliseconds, baseline; 86 6 12 milliseconds, 4 hours; 81 6 11 milliseconds, 8 hours; 83 6 6 milliseconds, 12 hours; 75 6 10 milliseconds, 24 hours; P , 0.05 for all, compared with baseline). Similar to the trend of ERP change, the HR was progressively decreased throughout the 24-hour pacing period (240 6 10 bpm, baseline; 214 6 19 bpm, 8 hours; 184 6 11 bpm, 24 hours; P , 0.05, compared with baseline) (Fig. 2C). For control animals, ERP and HR did not change over the 24-hour period, indicating that the animals remained stable during the entire experiment.
FIGURE 3. Changes in HRV parameters during the 24 hours of RAP. The LF (A) and HF (B) components and the LF/HF ratio (C) showed an initial increase with RAP, followed by a decrease. In contrast, SD of normalto-normal R-R interval (D) showed a progressive decrease over the 24-hour period. *P , 0.05 compared with baseline.
310
| www.jcvp.org
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Autonomic Nervous Remodeling in Paroxysmal AF
FIGURE 4. Typical examples of neural recordings of the SNA and VNA up to 24 hours. Note that the scale is different in each panel to optimize visualization. Each panel shows the recordings of the SNA and VNA. Frequency of neural activity was calculated by manually counting the clusters of firing. For example, the window denoted by the 2 arrows in the panel of 6H was counted as 1 cluster. At 24 hours, there were 12 clusters of VNA and 5 clusters of SNA over 10 seconds, translating to 60 and 30 clusters per minute.
Effects of RAP on HRV Analysis Figure 3 illustrates the changes in HRV parameters determined by time-domain and frequency-domain analysis. There was a significant increase in the HF component from 4 hours to 12 hours; it then attenuated during the next 12 hours (P , 0.05, compared with baseline; Fig. 3A). As the RAP continued, the LF component was significantly increased from 25 6 3 ms2 in the baseline state to 47 6 5 ms2 and 65 6 5 ms2 at the end of 4 hours and 8 hours, respectively (P , 0.05, Fig. 3B). Then, Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
LF decreased to 55 6 5 ms2 at the end of 12 hours and to 36 6 3 ms2 at the end of pacing. With RAP, LF/HF ratio increased during the first 12 hours (2.72 6 0.22, baseline; 4.23 6 0.37, 4 hours; 4.21 6 0.15, 8 hours; 3.38 6 0.28, 12 hours; P , 0.05 for all, compared with baseline; Fig. 3C). The LF/HF ratio reverted back to the baseline at 24 hours. SD of normal-to-normal R-R interval shortened progressively as the RAP duration continued to diminish (7.50 6 0.41 milliseconds at baseline; 6.30 6 0.53 milliseconds at www.jcvp.org |
311
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Zhang et al
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
FIGURE 5. Changes in SNA and VNA during the 24 hours of RAP. Frequency (A) and amplitude (B) of the SNA and VNA changes. C, The ratio of SNA/VNA amplitude over a 24-hour period. D, Correlation between the SNA and VNA. The clusters per minute (CPM) of VNA showed an initial increase with RAP, followed by a decrease. In contrast, the CPM of SNA gradually decreases and reaches a stable level at 6 hours. VNA has a higher CPM than SNA over the 24-hour period (A). The amplitude of VNA and SNA discharges progressively increased with RAP in the first 8 hours, and VNA remained stably elevated afterward; in contrast, SNA decreased at 24 hours. Of note, the SNA and VNA in the control group did not change over 24 hours. C, Increased VNA was associated with increased SNA from 2 to 12 hours. A good correlation between SNA and VNA was also found (r = 0.816 6 0.105, P , 0.001; D). *P , 0.05 compared with the baseline; #P , 0.05, compared with baseline in RAP group-VNA; ##P , 0.01, comparison between VNA and SNA.
4 hours; 5.65 6 0.54 milliseconds at 8 hours; 5.33 6 1.00 milliseconds at 12 hours; 3.77 6 0.43 milliseconds at 24 hours, P , 0.05, Fig. 3D). No time-dependent change was noted in the control group.
Effects of RAP on SNA and VNA Recording of the SNA and VNA was successful in all rabbits. In the baseline state, the VNA and SNA showed low-amplitude activities and increased progressively with RAP (Fig. 4). The VNA frequency and amplitude (Fig. 5A–B) showed an initial increase with RAP and plateaued between 12 and 24 hours. In contrast to VNA, the frequency of SNA did not change over time (Fig. 5A), but the amplitude of SNA increased from 9.89 6 2.48 mV to 21.41 6 3.84 mV at 4 hours and 89.36 6 8.98 at 8 hours and then remained stable from 8 to 12 hours. It was then followed by SNA withdrawal to 70.6 6 4.23 mV during 12–24 hours (Fig. 5B). The amplitude of VNA discharges progressively increased with RAP in the first 8 hours and remained stably elevated afterward. Of note, the SNA and VNA in the control group did not change over 24 hours. Figure 5C showed that increased VNA was associated with increased SNA from 2 to 12 hours. A good correlation between SNA and VNA was also found (r = 0.816 6 0.105, P , 0.001; Fig. 5D).
312
| www.jcvp.org
Histological Studies Immunohistochemical staining showed that GAP43-, ChAT-, and TH-immunoreactive neural elements were present in RA and LA tissues at 4, 8, 12, and 24 hours (Fig. 6A). In the baseline state, the RA had a slightly higher GAP43-positive and TH-positive nerve density than the LA, while the distribution of ChAT-positive nerves was reversed. The LA also contained more parasympathetic nerves than the RA. As RAP continued, GAP43-, ChAT-, and TH-positive nerve density progressively increased (P , 0.05; compared with control, Fig. 6B–D). Differential expression of TH-positive and ChAT-positive nerves between the RA and LA was also observed.
DISCUSSION Main Findings
To the best of our knowledge, this study is the first one to assess the interactions between electrical and autonomic remodeling during the first 24 hours after AF initiation. We provided experimental evidence for the presence of both electrical and autonomic remodeling. The former is characterized by progressive shortening of the ERP along with progressive slowing of the sinus rate (inverse of rate adaptation). The latter was characterized by increased VNA and SNA, corresponding changes in HRV parameters, and Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
Autonomic Nervous Remodeling in Paroxysmal AF
FIGURE 6. A, In the baseline state, the RA had a slightly higher GAP43-positive and TH-positive nerve density than the LA, while the distribution of ChAT-positive nerves was reversed. The LA also contained more parasympathetic nerves than the RA. As RAP continued, GAP43-, ChAT-, and TH-positive nerve density progressively increased compared with control. Differential expression of TH-positive and ChATpositive nerves between the RA and LA was also observed. Immunohistochemical staining for GAP43, ChAT, and TH antibodies in atrial tissue (dark brown stain) in control rabbits and in rabbits undergoing 4, 8, 12, and 24 hours of RAP. As RAP continued, GAP43 (B), TH(+) (C), and ChAT(+) (D) nerve density progressively increased. *P , 0.05 compared with the BS in the RA; #P , 0.05 compared with the BS in the LA; §P , 0.001, LA compared with the RA in the same group. Magnification, ·200.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.jcvp.org |
313
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Zhang et al
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
increased sympathetic and vagal innervation of the atria. Notably, ERP began to shorten and innervation began to increase after only 4 hours of RAP, indicating that both electrical and autonomic remodeling are necessary in maintaining AF in the first few hours after its initiation.
Electrical Remodeling to Perpetuate AF In 1995, 2 independent experimental studies reported that tachycardia-induced atrial electrical remodeling created a substrate for persistent AF, leading to the concept of AF begets AF.13,14 Both acute (hours) and chronic (weeks to months) RAP induced shortening of the atrial ERP and loss of rate adaptation.15–17 In this study, we found that with RAP, more pronounced ERP shortening was accompanied by slower HR, indicating that the physiological rate adaptation of the refractory period was inverted during the first 24 hours of simulated AF. Rapid atrial rate such as AF induced intracellular Ca2+ overload, followed by inactivation of L-type Ca++ channels. The reduced L-type Ca++ current shortens both the action potential duration and ERP.18 The transient outward potassium current (Ito) is reduced along with increased inward rectifier potassium current (IKI) and the acetylcholineactivated potassium current (IKACh). All of these factors contribute to the decreased action potential duration and rate adaptation changes.1 However, complete reversal of electrical remodeling can occur within a few days after restoration of sinus rhythm despite a prolonged period of AF (months).14,19,20 Without structural remodeling, which typically takes place after several weeks of persistent AF or RAP,18 electrical remodeling alone may not be able to maintain AF in the first 24 hours after AF was initiated.
Autonomic Remodeling to Perpetuate AF
Clinical and experimental studies have indicated fluctuation of the sympathetic and vagal tone preceding the onset of RAP.21,22 Vagal hyperactivity alone can shorten the ERP and increase ERP dispersion,23–25 promoting a substrate for reentry. Zhou et al proposed an “octopus hypothesis,” which states that the cardiac ANS is like an octopus. The head and tentacles of the octopus are the ganglionated plexi and autonomic nerves, respectively.26 This study is the first one to demonstrate specific patterns of autonomic nerve activity within the first 24 hours of AF. We found a progressive increase in the HF component and LF/HF ratio of HRV and an increase in SNA and VNA as well as the density of both the sympathetic and vagal nerves. These progressive changes were observed as early as 4 hours after RAP and continued until 12 hours. Between 12 and 24 hours, the nerve density continued to increase but the SNA and VNA decreased. At 24 hours, it reached a vagalpredominant state, evidence by a slower sinus rate (Fig. 2D) and a low SNA/VNA ratio (Fig. 5). We do not have experimental evidence to explain such dissociations between the continuously increased nerve density and attenuated nerve activity. We hypothesized that autonomic hyperinnervation may triggered an autonomic reflex to reduce the SNA and VNA as a protective mechanism. However, hyperinnervation and sprouting of the atrial autonomic nerves have set the stage for perpetuating AF beyond 24 hours. For example, the
314
| www.jcvp.org
hyperactive vagal activity at 24 hours can promote reentry by abbreviating the ERP and enhancing ERP dispersion. Stretch of the pulmonary veins or atrium can further abbreviate the ERP. Additional sympathetic stimulation by higher circulating catecholamine levels can easily facilitate the reinitiation and maintenance of AF beyond 24 hours. Recent experimental studies demonstrated that atrial nerve sprouting and sympathetic hyperinnervation provide a substrate for the development of AF.27–29 Atrial cholinergic innervation and peptidergic innervation also increased in animals undergoing long-term RAP and play an important role in the arrhythmogenesis of AF.29,30 The relative abundance of sympathetic hyperinnervation between the RA and LA often depends on the experimental model in which the electrical current delivered by the pacing lead is known to accelerate axonal regeneration31–33 and may result in nerve sprouting. In this study, sympathetic nerve sprouting appeared to be more pronounced in the RA than in the LA (Fig. 6). It is possible that this observation might be caused by rapid RA pacing. Regardless of the relative abundance of nerve sprouting between the RA and LA, heterogeneous atrial autonomic hyperinnervation is a common finding in patients with persistent or chronic AF.34–36 This study provides additional evidence to indicate the importance of the role of sympathetic and cholinergic hyperinnervation in the maintenance of AF.
Clinical Perspectives In this study, we focused on the mechanisms that maintain AF in the first 24 hours. Using a 24-hour RAP model in rabbits to simulate AF, we verified the presence of electrical remodeling as early as 4 hours. The SNA and VNA as well as the nerve density of the sympathetic and vagal nerves significantly began to progressively increase as early as 4 hours. So, we assume that autonomic remodeling, working synergistically with electrical remodeling, is important in maintaining AF and perpetuating AF beyond 24 hours. As the population ages, the incidence of AF is projected to sharply increase in the next 2–3 decades. As AF is a progressive disease, early interventions may prevent the development of irreversible changes such as fibrosis. This study calls for therapies designed specifically to inhibit autonomic remodeling in order to break the vicious cycle formed by electrical and autonomic remodeling. Early termination of AF will have a substantially positive impact on reducing the morbidities and mortality caused by more persistent forms of AF.
Study Limitations The stimulus strength of this study was relatively higher (twice diastolic threshold). In fact, the mean threshold is 0.35 6 0.10 mV. The higher stimulus strength might decrease the ERP in some certain. Because of the limitation of the neural recording device, VNA and SNA were not recorded continuously; at each time point, the SNA was recorded 3 minutes earlier than the VNA. Therefore, the exact temporal association between SNA and VNA was not studied. However, our experimental findings clearly illustrate the time course of the changes in SNA and VNA as well as the importance of progressive autonomic remodeling in AF maintenance. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol Volume 66, Number 3, September 2015
This study was conducted under general anesthesia that may introduce a confounding factor into the measurement of the ANS function, leading to the observation that the HRV data did not correlate well with the SNA and VNA after 12 hours. Also, ketamine might cause arrhythmias. However, all animals were under the same type of anesthesia, and the HR and HRV parameters were stable over 24 hours in the control group (Figs. 2–3), indicating that the changes in the ANS during the 24 hours of RAP are not merely artifacts from changes of ANS activity during anesthesia. To observe arrhythmia caused by ketamine, we have sham RAP group in which rabbits did not receive RAP for 24 hours to serve as controls; in this group, we have not observed any arrhythmia.
CONCLUSIONS Progressive autonomic remodeling manifested as autonomic nerve sprouting, sympathetic and vagal hyperinnervation, and a shift of the sympathetic/vagal balance in the first 24 hours of AF simulated by RAP. Working synergistically with electrical remodeling, autonomic remodeling may be a critical element in maintaining AF in the first 24 hours after its initiation. ACKNOWLEDGMENTS The authors are grateful to the Xinjiang Animal Research Center Facility staff Mei Ma and Mingjun Duan, Qing Wei, and Chun Zhang for their assistance. REFERENCES 1. Bosch RF, Zeng X, Grammer JB, et al. Ionic mechanisms of electrical remodeling in human atrial fibrillation. Cardiovasc Res. 1999;44: 121–131. 2. Verheule S, Wilson E, Everett T, et al. Alterations in atrial electrophysiology and tissue structure in a canine model of chronic atrial dilatation due to mitral regurgitation. Circulation. 2003;107:2615–2622. 3. Verheule S, Wilson E, Banthia S, et al. Direction-dependent conduction abnormalities in a canine model of atrial fibrillation due to chronic atrial dilatation. Am J Physiol Heart Circ Physiol. 2004;287:H634–H644. 4. Anderson EJ, Efird JT, Davies SW, et al. Monoamine oxidase is a major determinant of redox balance in human atrial myocardium and is associated with postoperative atrial fibrillation. J Am Heart Assoc. 2014;3:e000713. 5. Yu L, Scherlag BJ, Li S, et al. Low-level transcutaneous electrical stimulation of the auricular branch of the vagus nerve: a noninvasive approach to treat the initial phase of atrial fibrillation. Heart Rhythm. 2013;10:428–435. 6. Yu L, Scherlag BJ, Sha Y, et al. Interactions between atrial electrical remodeling and autonomic remodeling: how to break the vicious cycle. Heart Rhythm. 2012;9:804–809. 7. Xhyheri B, Manfrini O, Mazzolini M, et al. Heart rate variability today. Prog Cardiovasc Dis. 2012;55:321–331. 8. Camm AJ, Malik M, Bigger JT. Heart rate variability—standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation. 1996;93:1043–1065. 9. Jung BC, Dave AS, Tan AY, et al. Circadian variations of stellate ganglion nerve activity in ambulatory dogs. Heart Rhythm. 2006;3:78–85. 10. Piccirillo G, Magri D, Ogawa M, et al. Autonomic nervous system activity measured directly and QT interval variability in normal and pacing-induced tachycardia heart failure dogs. J Am Coll Cardiol. 2009;54:840–850. 11. Shen MJ, Choi EK, Tan AY, et al. Patterns of baseline autonomic nerve activity and the development of pacing-induced sustained atrial fibrillation. Heart Rhythm. 2011;8:583–589.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Autonomic Nervous Remodeling in Paroxysmal AF
12. Tan AY, Zhou S, Ogawa M, et al. Neural mechanisms of paroxysmal atrial fibrillation and paroxysmal atrial tachycardia in ambulatory canines. Circulation. 2008;118:916–925. 13. Morillo CA, Klein GJ, Jones DL, et al. Chronic rapid atrial pacing. Structural, functional, and electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation. 1995;91:1588–1595. 14. Wijffels MC, Kirchhof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation. 1995;92:1954–1968. 15. Lu Z, Scherlag BJ, Lin J, et al. Atrial fibrillation begets atrial fibrillation: autonomic mechanism for atrial electrical remodeling induced by shortterm rapid atrial pacing. Circ Arrhythm Electrophysiol. 2008;1:184–192. 16. Attuel P, Childers R, Cauchemez B, et al. Failure in the rate adaptation of the atrial refractory period: its relationship to vulnerability. Int J Cardiol. 1982;2:179–197. 17. Boutjdir M, Le Heuzey JY, Lavergne T, et al. Inhomogeneity of cellular refractoriness in human atrium: factor of arrhythmia? Pacing Clin Electrophysiol. 1986;9:1095–1100. 18. Gaspo R, Bosch RF, Talajic M, et al. Functional mechanisms underlying tachycardia-induced sustained atrial fibrillation in a chronic dog model. Circulation. 1997;96:4027–4035. 19. Resnekov L, McDonald L. Appraisal of electroconversion in treatment of cardiac dysrhythmias. Br Heart J. 1968;30:786–811. 20. Bjerkelund C, Orning OM. An evaluation of DC shock treatment of atrial arrhythmias. Acta Med Scand. 1968;184:481–491. 21. Bettoni M, Zimmermann M. Autonomic tone variations before the onset of paroxysmal atrial fibrillation. Circulation. 2002;105:2753–2759. 22. Fioranelli M, Piccoli M, Mileto GM, et al. Analysis of heart rate variability five minutes before the onset of paroxysmal atrial fibrillation. Pacing Clin Electrophysiol. 1999;22:743–749. 23. Goldberger AL, Pavelec RS. Vagally-mediated atrial fibrillation in dogs: conversion with bretylium tosylate. Int J Cardiol. 1986;13:47–55. 24. Oliveira M, da Silva MN, Geraldes V, et al. Acute vagal modulation of electrophysiology of the atrial and pulmonary veins increases vulnerability to atrial fibrillation. Exp Physiol. 2011;96:125–133. 25. Oliveira M, Silva MN, Geraldes V, et al. Effects of vagal stimulation on induction and termination of atrial fibrillation in an in vivo rabbit heart model. Rev Port Cardiol. 2010;29:375–389. 26. Zhou JSB, Edwards J, Jackman WM, et al. Gradients of atrial refractoriness and inducibility of atrial fibrillation due to stimulation of ganglionated plexi. J Cardiovasc Electrophysiol. 2007;18:83–90. 27. Jayachandran JV, Sih HJ, Winkle W, et al. Atrial fibrillation produced by prolonged rapid atrial pacing is associated with heterogeneous changes in atrial sympathetic innervation. Circulation. 2000;101:1185–1191. 28. Chang CM, Wu TJ, Zhou S, et al. Nerve sprouting and sympathetic hyperinnervation in a canine model of atrial fibrillation produced by prolonged right atrial pacing. Circulation. 2001;103:22–25. 29. Yu Y, Liu L, Jiang JY, et al. Parasympathetic and substance P-immunoreactive nerve denervation in atrial fibrillation models. Cardiovasc Pathol. 2012;21:39–45. 30. Qu X, Yu Y, Jiang J, et al. Variance of peptidic nerve innervation in a canine model of atrial fibrillation produced by prolonged atrial pacing. Pacing Clin Electrophysiol. 2008;31:207–213. 31. Nix WA, Hopf HC. Electrical stimulation of regenerating nerve and its effect on motor recovery. Brain Res. 1983;272:21–25. 32. Pockett S, Gavin RM. Acceleration of peripheral nerve regeneration after crush injury in rat. Neurosci Lett. 1985;59:221–224. 33. Manivannan S, Terakawa S. Rapid sprouting of filopodia in nerve terminals of chromaffin cells, PC12 cells, and dorsal root neurons induced by electrical stimulation. J Neurosci. 1994;14:5917–5928. 34. Gould PA, Yii M, McLean C, et al. Evidence for increased atrial sympathetic innervation in persistent human atrial fibrillation. Pacing Clin Electrophysiol. 2006;29:821–829. 35. Tan AY, Li HM, Wachsmann-Hogiu S, et al. Autonomic innervation and segmental muscular disconnections at the human pulmonary vein-atrial junction—implications for catheter ablation of atrial-pulmonary vein junction. J Am Coll Cardiol. 2006;48:132–143. 36. Li Y, Lu Z, Tang Q, et al. The increase in sympathetic nerve density in the atrium facilitates atrial fibrillation in patients with rheumatic heart disease. Int J Cardiol. 2013;165:174–178.
www.jcvp.org |
315
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
