Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Apr 8.
Published in final edited form as: Circulation. 2014 Jan 24;129(14):1472–1482. doi: 10.1161/CIRCULATIONAHA.113.004742

Dominant Frequency Increase Rate Predicts Transition from Paroxysmal to Long-Term Persistent Atrial Fibrillation

Raphael P Martins 1, Kuljeet Kaur 1, Elliot Hwang 1, Rafael J Ramirez 1, B Cicero Willis 1, David Filgueiras-Rama 1, Steven R Ennis 1, Yoshio Takemoto 1, Daniela Ponce-Balbuena 1, Manuel Zarzoso 1, Ryan P O’Connell 1, Hassan Musa 1, Guadalupe Guerrero-Serna 1, Uma Mahesh R Avula 1, Michael F Swartz 2, Sandesh Bhushal 3, Makarand Deo 3, Sandeep V Pandit 1, Omer Berenfeld 1, José Jalife 1
PMCID: PMC3981906  NIHMSID: NIHMS559237  PMID: 24463369

Abstract

Background

Little is known about the mechanisms underlying the transition from paroxysmal to persistent atrial fibrillation (AF). In an ovine model of long-standing persistent AF (LS-PAF) we tested the hypothesis that the rate of electrical and/or structural remodeling, assessed by dominant frequency (DF) changes, determines the time at which AF becomes persistent.

Methods and Results

Self-sustained AF was induced by atrial tachypacing. Seven sheep were sacrificed 11.5±2.3 days after the transition to persistent AF and without reversal to sinus rhythm (SR); 7 sheep were sacrificed after 341.3±16.7 days of LS-PAF. Seven sham-operated animals were in SR for 1 year. DF was monitored continuously in each group. RT-PCR, western blotting, patch-clamping and histological analyses were used to determine changes in functional ion channel expression and structural remodeling. Atrial dilatation, mitral valve regurgitation, myocyte hypertrophy, and atrial fibrosis occurred progressively and became statistically significant after the transition to persistent AF, with no evidence for left ventricular dysfunction. DF increased progressively during the paroxysmal-to-persistent AF transition and stabilized when AF became persistent. Importantly, the rate of DF increase (dDF/dt) correlated strongly with the time to persistent AF. Significant action potential duration (APD) abbreviation, secondary to functional ion channel protein expression changes (CaV1.2, NaV1.5 and KV4.2 decrease; Kir2.3 increase), was already present at the transition and persisted for one-year follow up.

Conclusions

In the sheep model of LS-PAF, the rate of DF increase predicts the time at which AF stabilizes and becomes persistent, reflecting changes in APD and densities of sodium, L-type calcium and inward rectifier currents.

Keywords: Electrical Remodeling, Structural Remodeling, Fibrosis, Ion Channels, Refractory Period

INTRODUCTION

Atrial fibrillation (AF) is the most common arrhythmia in clinical practice. The natural history of AF usually starts with paroxysmal episodes. Some patients suffer paroxysmal AF indefinitely, mainly under anti-arrhythmic therapy, but a large proportion progress to persistent AF. In these patients, progression from paroxysmal to persistent and permanent AF is likely to reflect progressive electrophysiological and/or structural remodeling in both atria, making the arrhythmia more stable and long-lasting.1 However, despite more than 100 years of research, fundamental mechanisms governing transition from paroxysmal to persistent and permanent forms are poorly understood, and prevention and treatment remain suboptimal.

Electrical remodeling, reflected by shortening of atrial refractoriness, is known to develop within the first few days of AF.2 While some ion channel changes associated with electrical remodeling have been described in animal models and humans,3–5 it is unclear yet how these changes integrate to stabilize AF. Structural remodeling and fibrosis might also contribute to intra-atrial conduction disturbances and increase susceptibility for AF, yet their role in progression from paroxysmal to persistent AF remains to be elucidated. Recently, we developed a clinically relevant ovine model of intermittent right atrial (RA) tachypacing and demonstrated that after the first AF episode, dominant frequency (DF) of both the RA and left atrium (LA) increased gradually during a 2-week period, after which DF remained stable during follow-up.6 Here we have extended our study to investigate systematically whether gradual DF increase can predict when AF becomes persistent. We have modified the model such that pacing stops temporarily when AF is initiated during paroxysmal episodes and permanently once AF is sustained without reverting to sinus rhythm (SR). We compared three groups: Sham-operated, Transition (>7 days of self-sustained AF without reversal to SR) and self-sustained Long-Standing Persistent AF (LS-PAF, > 1 year of AF without reversal to SR). We hypothesized that: 1) DF increase rate in paroxysmal AF predicts timing of transition from paroxysmal to persistent AF; and 2) electrophysiological remodeling occurs early in transition, whereas structural remodeling in the form of interstitial fibrosis appears more gradually and is belatedly manifest once self-sustained persistent AF has stabilized.

METHODS

An expanded methods section is available in the Online Supplement.

Pacemaker implantation

Procedures were approved by the University of Michigan Committee on Use and Care of Animals and complied with National Institutes of Health guidelines. Twenty-one 6–8 month-old sheep (≈40 kg) had a pacemaker implanted subcutaneously, with an atrial lead inserted into the RA appendage. In a subset of thirteen sheep (7 LS-PAF, 1 transition and 5 in SR), an implantable loop recorder (ILR) was placed subcutaneously on the left side of the sternum (Figure S1A).

Pacing protocol

After 10 days recovery, sheep were assigned to either Sham-operated or atrial tachypacing groups. Sham-operated animals (N=7) had pacemakers programmed in sensing-only mode (OAO). Automatic mode switch was enabled in atrial-tachypaced animals (Figure S1B). The pacemaker was programmed to induce AF by burst tachypacing (30-sec pacing, 20 Hz, twice diastolic threshold) followed by 10-sec sensing. Pacemakers resumed pacing only if AF stopped and SR was detected. The Holter capabilities of the device were used to record intra-cardiac electrograms (EGMs) to accurately confirm the occurrence of AF, generate histograms, and follow the evolution of AF from the first episode of paroxysmal AF to the eventual establishment of persistent AF. Persistent AF was defined as episodes lasting >7 days without reversal to SR. A subset of tachypaced animals was sacrificed after more than 7 days of self-sustained AF (Transition group, N=7). The remaining animals were sacrificed after one year of self-sustained AF (LS-PAF group, N=7). Pacemakers and ILRs were interrogated weekly.

Electrogram acquisition and processing

Persistence of SR was verified in sham-operated animals and pacemaker memories were checked to detect spontaneous AF episodes. Three recordings were obtained in tachypaced sheep during follow up: 1) RA-lead EGM; 2) ECG Standard lead I; and 3) ILR single lead recording (LA far-field signal).

Serum measurements

Procollagen III N-Terminal Propeptide (PIIINP) levels were measured by enzyme linked immunosorbant assay.

Echocardiography

LA and RA dimensions, severity of mitral regurgitation, left ventricular ejection fraction (LVEF), end-systolic and end-diastolic diameters were evaluated using echocardiography (Figures S2 and S3).7

Heart removal and cell dissociation

At termination of follow-up, hearts were removed and atria dissected (Figure S4) for myocyte isolation. patch-clamp recordings, western blotting, real-time PCR and histology (see Online Supplement).

Computer Simulations

We modified the Grandi-Pandit8 human atrial cell model to simulate action potential and propagation in 2D.

Statistical analyses

Normally distributed data are expressed as mean ± SEM. Normality of distributions was assessed using the Shapiro-Wilk test. A mixed regression model was applied to multiple group analyses and repeated measured data. Action potential durations (APD) and ionic current densities were compared using a two tailed unpaired Student’s-t tests. RT-PCR and Westerns blot data were analyzed using two-way ANOVA. A p<0.05 was considered statistically significant.

RESULTS

Sheep model of persistent AF

Of 21 implanted sheep one sham-operated animal was excluded and sacrificed prematurely due to severe symptomatic systemic infection. No atrial arrhythmias occurred in any sham-operated animals during follow-up. Also, no tachypaced animals developed signs of heart failure or stroke. Figure 1 summarizes the time-course of AF development. The representative 3D plot (Figure 1A) relates percentage of AF episodes in a given day (Y-axis) to duration of episodes (X-axis) and weeks of follow-up (Z-axis). The first AF episode occurred after a median time of 5.5 days after initiation of pacing (mean, 15.0±5.9 days; range, 0–62 days, Figure 1B). AF episodes were then paroxysmal (<7 days duration), reaching self-sustained persistent AF (>7 days without reversal to SR) after a median of 43.5 days (mean 73.2±23.0 days; range, 19 to 346 days). Once in persistent, there was no further tachypacing as AF was detected uninterruptedly. Sheep in Transition and LS-PAF were sacrificed 11.5±2.3 days and 341.3±16.7 days, respectively, after occurrence of self-sustained persistent AF (i.e. after the last occurrence of SR).

Figure 1.

Figure 1

Open in a new tab

Time-course of AF development. A: representative 3D plot of percentage of AF episodes in a given week (Y-axis) vs episode duration (X-axis) and weeks of follow-up after initiation of pacing (Z-axis). The first paroxysmal episode occurred 3 weeks after initiation of pacing. Duration of episodes progressively increased until persistent AF developed (week 12). B: summary of temporal measurements. AF: Atrial fibrillation; LS-PAF: Long-Standing Persistent AF.

Persistent AF leads to atrial dilatation

Echocardiographic findings (Table S1, Figures S2 and S3) revealed that LVEF was unchanged whereas RA and LA areas increased significantly in LS-PAF, compared with sham-operated and Transition groups (p<0.05, Figure S3). At last follow-up, LS-PAF animals showed significant mitral valve regurgitation (Figure S3B), yet LV end-diastolic volume, LV end-systolic volume or wall thickness were unchanged (data not shown), ruling out tachycardia-induced cardiomyopathy associated with AF. Although, compared to sham-operated animals, the dry weight of isolated atria in the transition group tended to be larger, only the atrial tissues from the LS-PAF group demonstrated a significant increase weight (Table S2).

Persistent AF leads to atrial myocyte hypertrophy

Mean LA and RA myocyte length and width, respectively, were similar for sham-operated animals (Figure S5). At transition, LA myocyte length and width increased significantly (p<0.001 and p<0.01, respectively); RA myocyte length did not change significantly (p=0.25) and a trend for wider cells was observed (p=0.08). At transition, LA cells were longer than RA cells (p<0.001), and after one year of AF, no further differences were observed for LA myocyte lengths or widths compared to transition. However, RA myocytes that initially did not exhibit significant changes at transition, showed a trend for longer cells and were significantly wider (p<0.001). In LS-PAF, LA myocytes were longer (p=0.002) and thinner (p=0.001) compared to RA.

AF leads to atrial myofibroblast activation and fibrosis in the absence of heart failure

AF-induced changes in the extracellular matrix were analyzed using histology, serum markers and molecular biology. There was a trend towards increased patchy fibrosis in RA, LA and PLA regions during AF progression, interstitial fibrosis increased in both LA (from 5.5±1.2 to 10.7±1.5%, p<0.05) and PLA (from 4.1±0.6 to 14.6±1.4%, p<0.001), particularly in LS-PAF (Figure 2A–B, Table S3). These data correlated with measurements of PIIINP, a serum marker for collagen synthesis, which increased progressively, reaching maximal levels in LS-PAF which was increased significantly from Sham-operated animals at a similar time point (p=0.001 vs. sham, Figure S6). As expected from PIIINP serum levels, tissue protein levels of collagen III, analyzed by western blot, increased significantly in both atria during LS-PAF (Figures 2C and D). A significant increase in atrial α-smooth muscle actin (α-SMA), a marker of myofibroblast activation,9 was seen in both atria in Transition, but tended to decrease toward control levels in LS-PAF.

Figure 2.

Figure 2

Open in a new tab

AF-induced changes in extracellular matrix. A: Mean±SEM values for patchy fibrosis (left) and interstitial fibrosis (right) in right atrium (RA), left atrium (LA) and posterior left atrium (PLA) of sham-operated (N=6), transition (N=7) and LS-PAF (N=7). Twenty pictures per slide were randomly selected and analyzed; *p<0.05; **p<0.001 vs. sham. B: Representative picrosirius red staining of PLA of sham-operated, transition and LS-PAF. C and D: Western blots of α-smooth muscle actin (SMA) and Collagen III (Col III) in LA and RA tissue homogenates relative to GAPDH. N=6 for each group. *p<0.05, **p<0.01 vs. sham. N = number of animals.

Electrophysiological remodeling is reflected by DF changes

During weekly interrogations, we investigated AF occurrence and recorded ongoing episodes. The DF of the first episode recorded from the RA lead was relatively slow at 7.5±0.1 Hz (range 6.5–8.25 Hz). Simultaneous DFs from the surface ECG and ILR after QRST subtraction were 7.7±0.2 Hz (range 6.5 to 9.25 Hz) and 9.0±0.1 Hz (range 8.9 – 9.4 Hz), respectively. Thus, as previously demonstrated,6 at the outset there was a significant DF difference between RA and LA (p<0.001, Figure 3). Thereafter, DF increased progressively in both atria. At both transition and LS-PAF, DFs recorded on the RA, surface ECG and LA were higher than during the first episode (p<0.001). However, in the 7 LS-PAF sheep, the last DFs recorded after 1 year of AF were not significantly different from prior corresponding values at transition. Thus, the major increase in DF occurred during paroxysmal AF and not during self-sustained LS-PAF. Additionally, while a significant LA-to-RA frequency gradient was present during the first episode, this gradient diminished at transition (p=0.06) and LS-persistent time points (p=0.1), likely reflecting remodeling of refractory periods in both atria. In any given animal, once respective maximum DF values were achieved, they remained relatively stable even after one year follow up; there was no significant difference between maximum DF at transition and at ~350 days.

Figure 3.

Figure 3

Open in a new tab

Dominant frequency increases in RA and LA (A) and surface ECG (B) during progression of AF. N=14 for RA, N=8 for LA. #p<0.001 for RA vs. LA, **p<0.001 vs. sham. N= number of animals.

The rate of DF increase predicts the onset of persistent AF

We analyzed several parameters to determine whether or not the time in paroxysmal AF and transition to self-sustained persistent AF could be predicted. We first surmised that a critical DF should be reached before self-sustained persistent AF developed, but the data did not support this hypothesis (Figure S7). Not only did maximal DF vary among animals, but the rate of DF increase during transition was also highly variable, ranging 0.003 to 0.15 Hz/day in the RA and 0.001 to 0.12 Hz/day in the LA. However, sheep that developed self-sustained persistent AF early, also had a steep slope of DF increase with time (dDF/dt), regardless of DF during the first episode, whereas those with a delayed onset of persistent AF had a shallower DF slope (Figure 4A). Thus we hypothesized that dDF/dt could predict when AF became persistent in each animal. Indeed, a strong nonlinear relationship was found between time to persistent AF onset and dDF/dt regardless of whether DF was determined in the RA, LA or surface ECG (R2= 0.87, 0.92 and 0.71, respectively, Figure 4B). The faster the DF increase, the quicker the animal developed self-sustained persistent AF. Furthermore, non-invasive measurement of dDF/dt (surface ECG lead I) correlated strongly with RA and LA dDF/dt (Figure S8).

Figure 4.

Figure 4

Open in a new tab

Rate of increase in DF during paroxysmal AF predicts transition to persistent AF. A: Representative graphs for three animals. Left, sheep with the highest dDF/dt (0.14 Hz/day, time to transition 19 days); middle, intermediate dDF/dt (0.03 Hz/day, time to transition 46 days); right, lowest dDF/dt (0.003 Hz/day, time to transition 346 days); left and right from transition group, middle from LS-PAF group. B: log-log plots of time from first episode to onset of self-sustained persistent AF versus dDF/dt for the RA (intracardiac electrode), LA (loop recorder) and ECG (surface Lead 1). Each point represents an animal. dDF/dt correlated with time to develop self-sustained persistent AF. N=14 for RA and ECG, N=8 for LA.

Cellular and ionic mechanisms of electrical remodeling

We conducted patch-clamp experiments to determine whether the gradual DF increase during transition reflected development of remodeling at the cellular level. Action potential duration at 90 percent repolarization (APD90) was significantly reduced in both RA and LA at transition and LS-PAF groups (Figure 5). Sheep from both groups tended to have more hyperpolarized resting membrane potentials than sham (p=NS, data not shown) for RA (−69.8±2.8 mV, −60.2±3.4 mV and −57.6±4.6 mV, respectively) and LA myocytes (−72.1±4.1 mV, −66.6±3.6 mV and −63.5±2.3 mV, respectively). Action potential (AP) upstroke velocity (dV/dtmax) also tended to be lower in myocytes from AF animals, while AP amplitudes did not change significantly (data not shown). Myocytes from animals in AF also showed a loss of rate-adaptation of APD (Figure 5B). Shortest pacing cycle length before AP alternans or failure to capture was significantly longer in sham than transition and LS-PAF groups (data not shown), as a consequence of APD and ERP shortening in both RA (345.7±37.5 ms, 165.7±62.6 ms and 203.3±26.5 ms, respectively, p<0.05 vs. sham) and LA (358.3±31.2 ms, 218.1±27.5 ms and 249.4±17.7 ms, respectively, p<0.05 vs. sham).

Figure 5.

Figure 5

Open in a new tab

APD and frequency dependence in myocytes from sham, transition, and persistent AF. A: Action potential duration (APD90 at 1Hz) is reduced in both atria at transition from paroxysmal to persistent AF. For RA: N=3/n=13 (sham), N=3/n=13 (transition), N=3/n=14 (LS-PAF); for LA: N=3/n=18 (sham), N=3/n=14 (transition), N=3/n=18 (LS-PAF). *p<0.05. Right: Representative LA APs are superimposed. B: Cycle length (CL) dependence of APD90. For RA: N=3/n=13 (sham), N=3/n=13 (transition), N=3/n=14 (LS-PAF); for LA: N=3/n=18 (sham), N=3/n=14 (transition), N=3/n=18 (LS-PAF). *p<0.05 Transition and LS-PAF vs. sham at 1000 ms CL. #p<0.05 Sham at 300ms vs. sham at v1000ms CL. N= number of animals; n= number of cells.

Next, we conducted Western blot analyses in the three groups on animals to test whether remodeling was related to altered intracellular calcium dysfunction. As illustrated in Figures S9 and S10 of the Online Supplement, while the Na+-Ca2+ exchanger was increased in the LA appendage, both total RyR2 and phosphorylated RyR2 proteins were decreased in the AF group, but the ratio of phosphorylated RyR2 to total RyR2 phosphorylation was unaffected. Accordingly, the transition from paroxysmal to persistent AF did not seem to depend on Ca2+ leak or delayed afterdepolarizations.10

We then focused on possible alterations in sarcolemmal ion channels that might contribute to AF-induced changes in APD and refractoriness. Peak inward sodium current (INa) was significantly reduced at the transition time-point by about 50% in LA myocytes compared to sham (Figure 6A) and about 30% in RA myocytes. For LS-PAF, peak INa was decreased in both LA and RA myocytes (p<0.001 vs. sham). Similarly, peak L-type calcium current (ICaL) was reduced in LA and RA at transition and LS-PAF (p<0.05, Figure 6B). Changes in INa and ICaL resulted from concomitant decreases in expression of Nav1.5 and Cav1.2 proteins and SCN5A and CACNA1C mRNA levels (Figure 6D–G; see Table S4 for primers used in RT-PCR).

Figure 6.

Figure 6

Open in a new tab

Sustained AF reduces functional expression of Na+ and L-type Ca2+ channels. A: Current-voltage relationships for INa in myocytes from LA (left) and RA (right). For LA: N=3/n=12 (sham), N=4/n=21 (transition), N=5/n=21 (LS-PAF); for RA: N=3/n=10 (sham), N=4/n=18 (transition), N=5/n=18 (LS-PAF). *p<0.05 vs. sham, # p<0.05 vs. transition. B: Current-voltage relationships for ICaL in myocytes from LA (left) and RA (right). For the LA: N=3/n=13 (sham), N=4/n=17 (transition), N=4/n=11 (LS-PAF); for the RA: N=3/n=12 (sham), N=4/n=16 (transition), N=4/n=14 (LS-PAF). *p<0.05 vs. sham. C: Representative traces for INa (upper) and ICaL (lower) in myocytes from LA of sham-operated and LS-PAF animal. D–E: Western blot analysis of NaV1.5 and CaV1.2 protein expression in LA tissue homogenates (D) and RA tissue homogenates (E). Top, Representative blots; bottom, Quantification of protein expression relative to GAPDH. N=6. F–G: Real time RT-PCR analysis of SNC5A and CACNA1C gene expression in tissue homogenates from LA (F) and RA (G); quantification of gene expression relative to GAPDH. N=6. **p< 0.01 vs. sham. N= number of animals; n= number of cells.

In contrast to INa and ICaL, the density of the inward rectifier potassium current (IK1) increased 2- to 3-fold at negative membrane voltages during the transition in both atria, and continued to increase for LS-PAF (p<0.05 vs. sham, Figure 7A). Since sheep atria predominantly express Kir2.3 channels,11 we measured Kir2.3 expression which was increased in LS-PAF animals (Figure 7B). There was no Kir2.3 increase in transition despite the larger current density compared to sham. The transient outward K+ current (Ito) decreased by about 85% by transition (Figure S11) and remained low in LS-PAF (p<0.001, Figure S11). For LS-PAF animals, Ito reduction could be explained by decreased Kv4.2 expression. However, reduced protein was not evidenced in the LA in transition animals, suggesting other mechanisms contributed to Ito decrease.12 Lastly, Kv11.1 protein expression remained unchanged (Figure S11C–D).

Figure 7.

Figure 7

Open in a new tab

Sustained AF increases functional expression of Kir2.3. A: Current-voltage relationships for IK1 in myocytes from LA (top) and RA (bottom). For LA: N=3/n=7 (sham), N=5/n=10 (transition), N=2/n=4 (LS-PAF); for RA: N=3/n=6 (sham), N=3/n=10 (transition), N=3/n=9 (LS-PAF). *p<0.05 vs. sham. B: Western blots for Kir2.3 in LA tissue homogenates. Top: representative blots of 2 different groups; bottom: quantification of protein expression relative to GAPDH. N=6. *p<0.05 vs. sham. N= number of animals; n= number of cells.

Can ionic current changes explain DF increase?

To address the question of whether differential changes in ion currents demonstrated above could explain DF increase during transition from paroxysmal to persistent AF we generated APs for control, paroxysmal, and transition AF conditions using the Grandi-Pandit human atrial AP model (Figure 8A, Table S5). The ionic changes for the transition AF were based on our experimental patch clamp recordings. We did not have ionic current recordings for paroxysmal AF. Therefore, to represent paroxysmal AF, we retained the ionic changes made in transition AF, but reduced the magnitude of ICaL by only 30% (Table S5), such that the simulated APD90 was shortened significantly by 17% in paroxysmal AF, compared to 51% in transition AF (Table S6).

Figure 8.

Figure 8

Open in a new tab

Simulations predict consequences of ion channel remodeling on rotor frequency. A: Action potential traces for sham, paroxysmal and transition AF predicted by experimentally derived ion channel changes (Figures 6–7). APD90 was abbreviated in both paroxysmal and transition AF compared to sham. Resting membrane potential was hyperpolarized −2 mV. B: Rotor in paroxysmal (left) had lower frequency than transition AF. C: Rotors in paroxysmal AF meandered considerably and eventually self-terminated upon collision with boundary. In transition AF, the rotor was stable, had higher frequency and persisted throughout the simulation.

We used a 2D sheet model of reentry to investigate whether AP differences between paroxysmal and transition AF simulations would explain the progressive DF increase demonstrated in vivo. Sustained functional reentry (rotor) dynamics showed differential properties. The rotor in paroxysmal AF (Figure 8B, left) was short lived, and exhibited low rotation frequency (5.0 Hz) and considerable meandering (Figure 8C, left), eventually self-terminating upon collision with boundary edges. In contrast, in the transition AF model, the rotor was stable and persisted throughout the length of the simulation (Figure 8B, right) with significantly less rotor meander (Figure 8C, right) and higher DF (7.67 Hz) compared to the transition case. When reduction in INa density was not incorporated, the DF increased only slightly to 8.67 Hz, but the rotor was unstable and eventually stopped.

We also further investigated the roles of individual ionic changes in a subset of simulations. Rotors were simulated in 2D sheets, when individual ionic currents were changed, compared to controls. As discussed in detail in the Online Supplement (Figures S12–S15), the simulation results confirmed that changes in IK1 and ICaL are key determinants of rotor acceleration in paroxysmal and transition AF.

Fast versus slow transition

To search for determinants of the rate of AF progression, we separated slow and fast progressing animals sacrificed at transition depending on the median time to progression (<45 days: 4 animals; >45 days: 3 animals). As discussed in detail in the Online Supplement (Figures S16 and S17), the major factor contributing to the larger dDF/dt in the fast transition animals was greater APD shortening secondary to ICaL reduction. On the other hand, the slow transition animals seemed to require an additional IK1 increase and greater structural remodeling.

DISCUSSION

The most important results of this study are: 1) Intermittent rapid tachypacing results in a progressive increase in DF during paroxysmal AF; 2) When DF stabilizes (dDF/dt>0), AF becomes persistent; 3) DF increase correlates strongly with time to persistent AF; 4) APD abbreviation, secondary to ion channel gene expression changes (NaV1.5 CaV1.2, and KV4.2 decrease; Kir2.3 increase), is already present or occurs rapidly during transition and explains the DF increase; 5) In the absence of LV dysfunction, there is a progressive increase in atrial dilatation, mitral valve regurgitation, myocyte hypertrophy, and atrial fibrosis, which became significant after DF had stabilized. Altogether, these results demonstrate that the rate of DF increase during transition predicts the time at which AF stabilizes and becomes persistent, reflecting changes in APD and densities of ICaL, IK1, INa and Ito. Thus, this is the first detailed characterization of the electrophysiological and structural remodeling involved in the transition from paroxysmal to persistent AF and self-sustained LS-PAF.

In-vivo changes in DF during AF

Various animal models and clinical studies have demonstrated the presence of a spatial distribution of DF during AF. Left-to-right frequency gradients were found in isolated sheep hearts, supporting the hypothesis that AF results from rapidly successive wavefronts emanating from fast sources localized in the LA.13 Similar gradients were confirmed in humans with AF,14, 15 but were shown to be present mainly during paroxysmal AF, but not always in persistent or permanent AF since longer periods in AF lead to a more complex remodeling making both atria suitable for harboring reentrant sources.14–16 Patients with persistent AF usually demonstrate higher DFs compared to paroxysmal AF.15 We recently demonstrated in the ovine model that intermittent tachypacing results in AF with a progressive increase in DF over a 2-week period after the first detected AF episode.6 There we used an algorithm of 30-seconds pacing followed by a 10-second blanking period, over a 22-week follow-up, whether or not AF was detected. Thus, continuous pacing likely induced a DF that was similar in all animals studied at that time.6 To increase the clinical relevance of our AF model, here we modified our pacing algorithm by transforming the 10-second blanking period into a real sensing period, consequently avoiding unnecessary pacing and generating a model of self-sustaining AF. Therefore, in this more realistic model of lone AF, there was a progressive increase in DF that was different for each animal, ranging from 4–50 weeks. The first AF episode was slow followed by a progressive and significant increase in DF during paroxysmal AF, reaching maximal value at the transition to persistent AF. Most important, DF did not increase further even after 1 year of self-sustained persistent AF.

Progressive DF increase during transition is a consequence of electrical remodeling. Sustained AF shortens APD and effective refractory period, decreasing wavelength and facilitating acceleration and stabilization of sustained reentry. The main determinants of APD shortening are the decrease in ICaL and increase IK1.17, 18 In a canine model of constant rapid atrial pacing, appreciable APD shortening occurred after 1 day of pacing, and was near-maximal after 7 days.19 At the electrophysiological level, reduced ICaL and Ito were observed, without significant change in IK1.19 However, human remodeling as a consequence of AF has been shown to be more complex, involving many changes in potassium currents like decreased of Ito15 and IKur5 and increased IKs5, IK1 and the constitutively-active IKACh.4 The contribution of these changes to APD shortening has been analyzed in computer modeling studies suggesting that IK1 increase is a predominant mechanism of APD shortening.20

Our results support previous numerical predictions. INa and ICaL decreases secondary to reductions in NaV1.5 and CaV1.5 expression, and IK1 density increases secondary to increased Kir2.3 expression were the major ionic changes observed at transition (remodeling during paroxysmal AF); they evolved parallel with DF increase, and changed negligibly throughout the one-year follow-up in self-sustained persistent AF. Upregulation of IK1 is known to enhance cardiac excitability through cell hyperpolarization and increased sodium channel availability, and to increase frequency and stability of rotors driving AF and VF.21 Our computer simulations reaffirmed these results since rotors were found to be more stable, less meandering and persistent for transition compared to paroxysmal time-points (7.7 vs. 5.0 Hz, respectively). Similarly, INa reduction, despite reducing excitability slightly, also contributed to rotor stabilization, as shown by the fact that rotors generated in the absence of INa reduction exhibited considerable meandering leading to their eventual annihilation. Most important, however, our numerical results predict that downregulation of ICaL and upregulation of IK1 to be the most important contributors to the increased DF and rotor stabilization, particularly during transition AF.

Structural remodeling and the link to electrical remodeling

We observed significant structural remodeling in this model of lone, persistent AF. Myocyte hypertrophy and atrial dilatation increased progressively among sham-operated sheep and sheep at transition or LS-PAF, which suggests there might be a connection between atrial enlargement and DF increase. Additionally, PIIINP, a serum marker of interstitial fibrosis, increased gradually from the first episode through transition into the last follow-up. Collagen protein levels and interstitial fibrosis also tended to increase during transition and became frankly manifest once self-sustained LS-PAF developed. Notably, a significant increase in atrial α-SMA protein, a marker of myofibroblast activation,9 was seen in both atria in transition, but this protein tended to decrease toward control levels in LS-PAF animals. These changes might reflect processes similar to those described for myocardial infarction: a myofibroblast proliferative phase gives way to a maturation phase in which the cellularity of tissue decreases and local extracellular matrix is cross-linked forming a collagen-based fibrotic scar.22 Therefore, the increase in α-SMA protein may reflect increased atrial myofibroblast activation and proliferation induced by atrial tachypacing induced upregulation of transient receptor potential canonical-3 (TRPC3) channels,23 which may have contributed not only to development of fibrosis, but also to electrical remodeling through release of profibrotic cytokines. One such cytokine is platelet-derived growth factor (PDGF), which, as recently demonstrated, can reduce atrial myocyte APD and ICaL in ways that resemble the effects of persistent AF as shown in this study.24 Finally, we found no change or downregulation of most calcium-handling proteins and no change in the phosphorylated RyR2/total RyR2 ratio, which makes it unlikely that, in this animal model of lone AF, the rate of change of AF-induced differential ion channel expression depends on sustained AF-induced intracellular Ca2+-handling remodeling or increased SR Ca2+ leak. However, it is still possible that Ca2+ loading produced by the high AF frequency activated the Ca2+-dependent calmodulin-calcineurin-NFAT system to cause transcriptional downregulation of ICaL.25

Predicting transition from paroxysmal to persistent AF

Mid- to long-term follow-up studies in patients have shown that the rate of progression from paroxysmal to persistent or permanent AF ranges between 14.6% and 35.3% during 1 to 12 years.26–29 A 30-year follow-up study reported a cumulative probability risk of progression of 29% (95% CI, 16 to 42%) with most transitions occurring within 15 years after diagnosis. Many independent clinical, echocardiographic and electrocardiographic predictors of transition have been described including age,1, 27–29 hypertension, previous ischemic attack or stroke, chronic obstructive pulmonary disease,1 presence of cardiomyopathy/heart failure,1, 26–28 atrial enlargement,28 valvular diseases28, filtered P-wave duration ≥150 ms, or P-wave dispersion on ECG.29 A report using pacemaker memory interrogation found 24% risk of transition at a mean follow-up of 5 months.26 Interestingly, slope of change in AT/AF burden as a function of time was examined. Patients developing sustained AF showed increase of AF burden over time, compared with patients remaining in paroxysmal AF who demonstrated no progressive change in AF duration (14 sec/day vs. 0 sec/day, p<0.001). Similarly, in our study, we observe a very different rate of increase of dDF/dt and episode durations between animals quickly developing persistent AF and those remaining in paroxysmal AF for a long period before transition. Both LA and RA dDF/dt predicted the transition to persistent AF. This is not surprising since DF increased in both atria until AF became persistent (Figure 3). Thus ionic remodeling occurred in both atria, though somewhat more intensely in the LA than RA, particularly for ICaL (Figure 6B) and IK1 (Figure 7), both of which contributed to the increased DF and rotor stabilization (Figures S12 and S13). Finally, the possible contribution of cytokines released from ion-channel mediated activation30 and differentiation of fibroblasts into hypersecretory myofibroblasts as a common link to structural and electrical remodeling in the fibrillating atria is likely to be important and deserve further exploration.

Limitations

We used a pacing induced self-sustained AF model in which continued periods of SR were not allowed since pacing was resumed as soon as AF terminated (5–10 seconds of SR between spontaneous termination of AF, detection of SR and initiation of the following pacing run). Thus, it is unknown whether the time from paroxysmal to persistent AF would be predictable in other animal models or humans. In addition, while this model enabled us to investigate the consequences of sustained AF in the absence of other co-morbidities, we are well aware that the majority of human AF cases associate with cardiovascular disease. Nevertheless, our results show that the time to AF persistence is predicted by the rate of change of DF and explained primarily by downregulation of ICaL and upregulation of IK1. The results also suggest that the wide range of dDF/dt values and time to persistent AF might be explained by temporal differences in the remodeling of these two channels in animals exhibiting slow-versus-fast progression to LS-PAF, which requires further study. Other currents may be important, including IK118 and IKACh, whose constitutively active isoform is upregulated.4, 31 Activation of IKACh with adenosine accelerates DF in patients with paroxysmal AF16. Therefore studies should be performed to analyze the time-course of IKACh changes during AF progression. In our study Ito decrease and IK1 increase in transition did not correlate with protein expression, suggesting other genes encoding ion channel alpha- or beta-subunits might be important. Alternatively, open-channel probability in the case of IK131 might help explain the changes we have observed. Similarly, while our analysis of Ca2+-handling proteins suggested that the transition from paroxysmal to persistent AF did not depend on Ca2+ leak or DADs, more extensive studies involving intracellular Ca2+ measurements will be required for definite proof.25 Lastly, we decided not to model LS-PAF since no major electrophysiological changes occurred during the one-year follow-up. However, additional structural remodeling after the transition (e.g., atrial dilatation, fibrosis, myocyte hypertrophy) are likely to affect DF, but were not incorporated in the simulations.

CLINICAL IMPLICATIONS

Our model of intermittent atrial tachypacing resembles human AF in the absence of co-morbidities. Similar to humans, AF in sheep follows a very heterogeneous temporal progression to persistent AF, ranging from 4–50 weeks. Therefore, analyzing dDF/dt by surface ECG in human AF could help stratify paroxysmal AF patients depending on risk of progression and be of great help in guiding physicians to individualize therapy. Patients exhibiting steep dDF/dt increases during follow-up could be rapidly referred to an EP lab for ablation before persistent AF develops, which would reduce ablation time and number of procedures.32 Conversely, patients with low dDF/dt could be considered at low risk for progression and be treated longer with anti-arrhythmic drugs before being referred to an EP lab.

Supplementary Material

supplemental material

Potential Clinical Impact.

We have used a model of intermittent atrial tachypacing that resembles closely the situation in humans with AF in the absence of co-morbidities like left ventricular dysfunction and heart failure. We demonstrate that predicting transition from paroxysmal to persistent AF is feasible. Similar to human AF, we found a very heterogeneous temporal progression to persistent AF among animals, ranging from 4–50 weeks. Similar findings in human AF would help stratify paroxysmal AF patients depending on risk of progression. Patients exhibiting steep dDF/dt increases during follow-up could be rapidly referred to an EP lab for ablation before persistent AF develops, which would reduce ablation time and number of procedures. Conversely, patients with low dDF/dt could be considered at low risk for progression and be treated longer with anti-arrhythmic drugs before being referred to an EP lab. We showed also that dDF/dt analyzed by surface ECG (lead I) correlated well with RA and LA dDF/dt. This would obviate invasive analysis for stratification and be of great help in guiding physicians to individualize therapy.

Acknowledgments

Authors would like to thank Felipe Atienza, Justus M.B. Anumonwo, Mario San Martin-Gomez, and Jérôme Kalifa for their helpful suggestions.

Funding Sources: This work was supported in part by, the National Heart, Lung, and Blood Institute Grants (P01-HL039707 and P01-HL087226 to JJ and OB, R01-HL118304 to OB), the Leducq Foundation and CNIC (JJ and OB) the Fédération Française de Cardiologie (RPM), Spanish Society of Cardiology Fellowship, and Alfonso Martín Escudero Foundation (DFR).

Footnotes

Conflict of Interest Disclosures: Dr. Jalife is on the Scientific Advisory Board of avertAF. Dr. Berenfeld is the Scientific Officer of Rhythm Solutions, Inc.

References

  • 1.de Vos CB, Pisters R, Nieuwlaat R, Prins MH, Tieleman RG, Coelen RJ, van den Heijkant AC, Allessie MA, Crijns HJ. Progression from paroxysmal to persistent atrial fibrillation clinical correlates and prognosis. J Am Coll Cardiol. 2010;55:725–731. doi: 10.1016/j.jacc.2009.11.040. [DOI] [PubMed] [Google Scholar]
  • 2.Allessie M, Ausma J, Schotten U. Electrical, contractile and structural remodeling during atrial fibrillation. Cardiovasc Res. 2002;54:230–246. doi: 10.1016/s0008-6363(02)00258-4. [DOI] [PubMed] [Google Scholar]
  • 3.Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM. Atrial l-type ca2+ currents and human atrial fibrillation. Circ Res. 1999;85:428–436. doi: 10.1161/01.res.85.5.428. [DOI] [PubMed] [Google Scholar]
  • 4.Voigt N, Trausch A, Knaut M, Matschke K, Varro A, Van Wagoner DR, Nattel S, Ravens U, Dobrev D. Left-to-right atrial inward rectifier potassium current gradients in patients with paroxysmal versus chronic atrial fibrillation. Circ Arrhythm Electrophysiol. 2010;3:472–480. doi: 10.1161/CIRCEP.110.954636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Caballero R, de la Fuente MG, Gomez R, Barana A, Amoros I, Dolz-Gaiton P, Osuna L, Almendral J, Atienza F, Fernandez-Aviles F, Pita A, Rodriguez-Roda J, Pinto A, Tamargo J, Delpon E. In humans, chronic atrial fibrillation decreases the transient outward current and ultrarapid component of the delayed rectifier current differentially on each atria and increases the slow component of the delayed rectifier current in both. J Am Coll Cardiol. 2010;55:2346–2354. doi: 10.1016/j.jacc.2010.02.028. [DOI] [PubMed] [Google Scholar]
  • 6.Filgueiras-Rama D, Price NF, Martins RP, Yamazaki M, Avula UM, Kaur K, Kalifa J, Ennis SR, Hwang E, Devabhaktuni V, Jalife J, Berenfeld O. Long-term frequency gradients during persistent atrial fibrillation in sheep are associated with stable sources in the left atrium. Circ Arrhythm Electrophysiol. 2012;5:1160–1167. doi: 10.1161/CIRCEP.111.969519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Gottdiener JS, Bednarz J, Devereux R, Gardin J, Klein A, Manning WJ, Morehead A, Kitzman D, Oh J, Quinones M, Schiller NB, Stein JH, Weissman NJ. American society of echocardiography recommendations for use of echocardiography in clinical trials. J Am Soc Echocardiogr. 2004;17:1086–1119. doi: 10.1016/j.echo.2004.07.013. [DOI] [PubMed] [Google Scholar]
  • 8.Grandi E, Pandit SV, Voigt N, Workman AJ, Dobrev D, Jalife J, Bers DM. Human atrial action potential and ca2+ model: Sinus rhythm and chronic atrial fibrillation. Circ Res. 2011;109:1055–1066. doi: 10.1161/CIRCRESAHA.111.253955. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Frangogiannis NG, Michael LH, Entman ML. Myofibroblasts in reperfused myocardial infarcts express the embryonic form of smooth muscle myosin heavy chain (smemb) Cardiovasc Res. 2000;48:89–100. doi: 10.1016/s0008-6363(00)00158-9. [DOI] [PubMed] [Google Scholar]
  • 10.Voigt N, Li N, Wang Q, Wang W, Trafford AW, Abu-Taha I, Sun Q, Wieland T, Ravens U, Nattel S, Wehrens XH, Dobrev D. Enhanced sarcoplasmic reticulum ca2+ leak and increased na+-ca2+ exchanger function underlie delayed afterdepolarizations in patients with chronic atrial fibrillation. Circulation. 2012;125:2059–2070. doi: 10.1161/CIRCULATIONAHA.111.067306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Dhamoon AS, Pandit SV, Sarmast F, Parisian KR, Guha P, Li Y, Bagwe S, Taffet SM, Anumonwo JM. Unique kir2.X properties determine regional and species differences in the cardiac inward rectifier k+ current. Circ Res. 2004;94:1332–1339. doi: 10.1161/01.RES.0000128408.66946.67. [DOI] [PubMed] [Google Scholar]
  • 12.Tessier S, Karczewski P, Krause EG, Pansard Y, Acar C, Lang-Lazdunski M, Mercadier JJ, Hatem SN. Regulation of the transient outward k(+) current by ca(2+)/calmodulin-dependent protein kinases ii in human atrial myocytes. Circ Res. 1999;85:810–819. doi: 10.1161/01.res.85.9.810. [DOI] [PubMed] [Google Scholar]
  • 13.Mansour M, Mandapati R, Berenfeld O, Chen J, Samie FH, Jalife J. Left-to-right gradient of atrial frequencies during acute atrial fibrillation in the isolated sheep heart. Circulation. 2001;103:2631–2636. doi: 10.1161/01.cir.103.21.2631. [DOI] [PubMed] [Google Scholar]
  • 14.Lazar S, Dixit S, Marchlinski FE, Callans DJ, Gerstenfeld EP. Presence of left-to-right atrial frequency gradient in paroxysmal but not persistent atrial fibrillation in humans. Circulation. 2004;110:3181–3186. doi: 10.1161/01.CIR.0000147279.91094.5E. [DOI] [PubMed] [Google Scholar]
  • 15.Sanders P, Berenfeld O, Hocini M, Jais P, Vaidyanathan R, Hsu LF, Garrigue S, Takahashi Y, Rotter M, Sacher F, Scavee C, Ploutz-Snyder R, Jalife J, Haissaguerre M. Spectral analysis identifies sites of high-frequency activity maintaining atrial fibrillation in humans. Circulation. 2005;112:789–797. doi: 10.1161/CIRCULATIONAHA.104.517011. [DOI] [PubMed] [Google Scholar]
  • 16.Atienza F, Almendral J, Moreno J, Vaidyanathan R, Talkachou A, Kalifa J, Arenal A, Villacastin JP, Torrecilla EG, Sanchez A, Ploutz-Snyder R, Jalife J, Berenfeld O. Activation of inward rectifier potassium channels accelerates atrial fibrillation in humans: Evidence for a reentrant mechanism. Circulation. 2006;114:2434–2442. doi: 10.1161/CIRCULATIONAHA.106.633735. [DOI] [PubMed] [Google Scholar]
  • 17.Bosch RF, Scherer CR, Rub N, Wohrl S, Steinmeyer K, Haase H, Busch AE, Seipel L, Kuhlkamp V. Molecular mechanisms of early electrical remodeling: Transcriptional downregulation of ion channel subunits reduces i(ca,l) and i(to) in rapid atrial pacing in rabbits. J Am Coll Cardiol. 2003;41:858–869. doi: 10.1016/s0735-1097(02)02922-4. [DOI] [PubMed] [Google Scholar]
  • 18.Dobrev D, Graf E, Wettwer E, Himmel HM, Hala O, Doerfel C, Christ T, Schuler S, Ravens U. Molecular basis of downregulation of g-protein-coupled inward rectifying k(+) current (i(k,ach) in chronic human atrial fibrillation: Decrease in girk4 mrna correlates with reduced i(k,ach) and muscarinic receptor-mediated shortening of action potentials. Circulation. 2001;104:2551–2557. doi: 10.1161/hc4601.099466. [DOI] [PubMed] [Google Scholar]
  • 19.Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S. Ionic remodeling underlying action potential changes in a canine model of atrial fibrillation. Circ Res. 1997;81:512–525. doi: 10.1161/01.res.81.4.512. [DOI] [PubMed] [Google Scholar]
  • 20.Zhang H, Garratt CJ, Zhu J, Holden AV. Role of up-regulation of ik1 in action potential shortening associated with atrial fibrillation in humans. Cardiovasc Res. 2005;66:493–502. doi: 10.1016/j.cardiores.2005.01.020. [DOI] [PubMed] [Google Scholar]
  • 21.Pandit SV, Jalife J. Rotors and the dynamics of cardiac fibrillation. Circ Res. 2013;112:849–862. doi: 10.1161/CIRCRESAHA.111.300158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Dobaczewski M, Gonzalez-Quesada C, Frangogiannis NG. The extracellular matrix as a modulator of the inflammatory and reparative response following myocardial infarction. J Mol Cell Cardiol. 2010;48:504–511. doi: 10.1016/j.yjmcc.2009.07.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Harada M, Luo X, Qi XY, Tadevosyan A, Maguy A, Ordog B, Ledoux J, Kato T, Naud P, Voigt N, Shi Y, Kamiya K, Murohara T, Kodama I, Tardif JC, Schotten U, Van Wagoner DR, Dobrev D, Nattel S. Transient receptor potential canonical-3 channel-dependent fibroblast regulation in atrial fibrillation. Circulation. 2012;126:2051–2064. doi: 10.1161/CIRCULATIONAHA.112.121830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Musa H, Kaur K, O'Connell R, Klos M, Guerrero-Serna G, Avula UM, Herron TJ, Kalifa J, Anumonwo JM, Jalife J. Inhibition of platelet-derived growth factor-ab signaling prevents electromechanical remodeling of adult atrial myocytes that contact myofibroblasts. Heart Rhythm. 2013;10:1044–1051. doi: 10.1016/j.hrthm.2013.03.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Qi XY, Yeh YH, Xiao L, Burstein B, Maguy A, Chartier D, Villeneuve LR, Brundel BJ, Dobrev D, Nattel S. Cellular signaling underlying atrial tachycardia remodeling of l-type calcium current. Circ Res. 2008;103:845–854. doi: 10.1161/CIRCRESAHA.108.175463. [DOI] [PubMed] [Google Scholar]
  • 26.Saksena S, Hettrick DA, Koehler JL, Grammatico A, Padeletti L. Progression of paroxysmal atrial fibrillation to persistent atrial fibrillation in patients with bradyarrhythmias. Am Heart J. 2007;154:884–892. doi: 10.1016/j.ahj.2007.06.045. [DOI] [PubMed] [Google Scholar]
  • 27.Potpara TS, Stankovic GR, Beleslin BD, Polovina MM, Marinkovic JM, Ostojic MC, Lip GY. A 12-year follow-up study of patients with newly diagnosed lone atrial fibrillation: Implications of arrhythmia progression on prognosis: The belgrade atrial fibrillation study. Chest. 2012;141:339–347. doi: 10.1378/chest.11-0340. [DOI] [PubMed] [Google Scholar]
  • 28.Kerr CR, Humphries KH, Talajic M, Klein GJ, Connolly SJ, Green M, Boone J, Sheldon R, Dorian P, Newman D. Progression to chronic atrial fibrillation after the initial diagnosis of paroxysmal atrial fibrillation: Results from the canadian registry of atrial fibrillation. Am Heart J. 2005;149:489–496. doi: 10.1016/j.ahj.2004.09.053. [DOI] [PubMed] [Google Scholar]
  • 29.Koide Y, Yotsukura M, Ando H, Aoki S, Suzuki T, Sakata K, Ootomo E, Yoshino H. Usefulness of p-wave dispersion in standard twelve-lead electrocardiography to predict transition from paroxysmal to persistent atrial fibrillation. Am J Cardiol. 2008;102:573–577. doi: 10.1016/j.amjcard.2008.04.065. [DOI] [PubMed] [Google Scholar]
  • 30.Yue L, Xie J, Nattel S. Molecular determinants of cardiac fibroblast electrical function and therapeutic implications for atrial fibrillation. Cardiovasc Res. 2011;89:744–753. doi: 10.1093/cvr/cvq329. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Dobrev D, Friedrich A, Voigt N, Jost N, Wettwer E, Christ T, Knaut M, Ravens U. The g protein-gated potassium current i(k,ach) is constitutively active in patients with chronic atrial fibrillation. Circulation. 2005;112:3697–3706. doi: 10.1161/CIRCULATIONAHA.105.575332. [DOI] [PubMed] [Google Scholar]
  • 32.Roten L, Derval N, Pascale P, Scherr D, Komatsu Y, Shah A, Ramoul K, Denis A, Sacher F, Hocini M, Haissaguerre M, Jais P. Current hot potatoes in atrial fibrillation ablation. Curr Cardiol Rev. 2012;8:327–346. doi: 10.2174/157340312803760802. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental material
NIHMS559237-supplement-supplemental_material.pdf (1.8MB, pdf)

RESOURCES