Shorter Premature Atrial Complex Coupling Interval Leads to Mechanical Dysfunction, Fibrosis, and AF in Swine

Abstract

BACKGROUND

We have previously shown that dyssynchronous premature atrial complexes (PACs) from the lateral left atrium (LA) lead to greater atrial mechanical dysfunction, remodeling, and sustained atrial fibrillation (AF) than synchronous PACs from the interatrial septum. However, the impact of PAC coupling interval (CI) on atrial remodeling is unclear.

OBJECTIVES

This study sought to explore the effect of PAC CI on atrial mechanics and remodeling in the swine model.

METHODS

A 2-phase in vivo study was conducted. In the phase 1 acute study, 5 swine underwent acute invasive hemodynamics and echocardiography while delivering single-paced atrial extrastimuli with CIs varying from 450 ms to 200 ms. Peak LA longitudinal strain and intra-LA dyssynchrony were assessed with 2-dimensional strain echocardiography while LA and aortic pressure were directly measured. In the phase 2 chronic study, a group exposed to paced bigeminy from the lateral LA for 16 weeks with a short CI of 250 ms (Short-PAC, n = 10) was compared with groups with PACs at a long CI of 400 ms (Long-PAC, n = 5) and a nonpaced control group (CTRL, n = 10). Detailed electrophysiology and echocardiography studies were performed with histologic quantification of LA fibrosis at baseline and prior to sacrifice.

RESULTS

Phase 1 revealed that as PAC CI shortened, peak LA strain decreased (P = 0.003) and LA dyssynchrony increased (P < 0.001). Phase 2 showed that after 16 weeks of PACs, the Short-PAC group had greater LA dilation (terminal baseline: 5.9 ± 1.2 cm² vs Long-PAC 3.9 ± 0.5 cm² vs CTRL 0.9 ± 0.4 cm²; P < 0.001) and reduced peak LA strain during sinus rhythm (terminal baseline: −17.3% ± 3.2% vs Long-PAC −12.1% ± 2.1% vs CTRL −0.7% ± 4.2%; P < 0.001). The short-PAC group had a more LA fibrosis (8.6% ± 1.0% vs Long-PAC 6.8% ± 1.0% vs CTRL 4.0% ± 1.5%; P < 0.001) and higher AF inducibility (terminal baseline: 49.3% ± 13.0% vs Long-PAC 29.0% ± 6.4% vs CTRL 2.2% ± 16.2%; P < 0.001) than the other groups.

CONCLUSIONS

In this swine model, shorter PAC CI led to increased acute atrial mechanical dysfunction and dyssynchrony. Chronically, short-CI PACs led to greater atrial fibrosis and induced AF, suggesting that frequent, short-coupled PACs pose the highest risk for LA myopathy and AF. These insights underscore the importance of understanding the impact of PAC characteristics on atrial remodeling and arrhythmogenesis.

CENTRAL ILLUSTRATION

Frequent, Short-Coupled PACs Pose Higher Risk for Developing LA Myopathy and AF Than Long-Coupled PACs

(Top) Acute effect of short coupling interval (CI) premature atrial complexes (PACs): PACs with progressively shorter CIs resulted in lower peak left atrial (LA) strain and greater LA dyssynchrony and dyssynergy. Moreover, there was a greater increase in LA pressures at the time of a waves between pre-PAC and post-PAC at shorter CIs than at longer CIs. (Bottom) Chronic effect of short-CI PACs: prolonged exposure to frequent short-CI PACs over 16 weeks led to a greater degree of structural fibrotic remodeling and induced AF compared with long-CI PACs. AF = atrial fibrillation; ANOVA = analysis of variance; CTRL = control.

Premature atrial complexes (PACs) are frequently encountered in clinical practice and have largely been considered benign. However, emerging evidence suggests that frequent PACs may contribute to the development of atrial fibrillation (AF)^1–8 and adverse atrial remodeling,^9–11 akin to the relationship observed between premature ventricular complexes and ventricular cardiomyopathy.^12–14 In a previous study utilizing a swine model, we demonstrated that frequent short-coupled PACs can induce atrial structural remodeling and enhance susceptibility to AF, highlighting the potential significance of PACs in the pathogenesis of AF.¹⁵ In addition, we found that despite an identical PAC burden, more dyssynchronous PACs led to more atrial remodeling, fibrosis and AF, arguing that the atrial dyssynchrony was a major contributor to the ensuing atrial myopathy.¹⁵

The prior studies were all performed with an identical, short coupling interval (CI) of 250 ms, which resulted in nonconducted PACs, to eliminate the effect of PAC coupling and ventricular irregularity.¹⁵ Therefore, the objective of the current study was to investigate the relationship of PAC CI to atrial remodeling and AF. We hypothesized that frequent PACs with short CIs would result in more pronounced atrial dyssynchrony and atrial remodeling compared with PACs with long CIs or control animals without PACs.

METHODS

This was a 2-phase, in vivo study. In the first acute phase, 5 female Yucatan mini-swine underwent acute invasive hemodynamic and echocardiographic study to measure simultaneous left atrial (LA) and aortic pressure and LA mechanics at various PAC CIs. In the second chronic phase, mini-swine underwent a chronic study to investigate the influence of short vs long PAC CIs on the development of atrial remodeling and AF. All paced PACs during the phase 1 and 2 studies were produced from the lateral LA via the coronary sinus, as our prior study demonstrated that this led to a more pronounced atrial myopathy. Transthoracic speckle-tracking echocardiography in the phase 1 and 2 studies was performed by an experienced echocardiography technician (D.B.). The study was approved and overseen by the Laboratory Animal Resource Center at the University of California-San Francisco.

PHASE 1: ACUTE STUDY.

Following an overnight fast, anesthesia was induced with an intramuscular injection of ketamine and acepromazine and maintained by inhalation of 1% to 5% isoflurane, with each swine mechanically ventilated using 100% oxygen. Femoral venous access was obtained via percutaneous puncture using the modified Seldinger technique with ultrasound guidance. A transseptal puncture was undertaken guided by intracardiac echocardiography and fluoroscopy using a deflectable sheath (Agilis; Abbott Technologies) and BRK needle. Systemic heparinization was maintained throughout left atrial access. After introducing a deflectable sheath into the LA, continuous LA pressure was recorded during sinus rhythm and paced PACs delivered at different CIs (Supplemental Figure 1A). Aortic pressure was also recorded from the femoral arterial sheath. Paced PACs were produced by a steerable mapping catheter placed at the distal coronary sinus through a femoral venous sheath. Single-paced atrial extrastimuli at twice the diastolic threshold were delivered every 10 sensed beats from the pacing catheter to mimic spontaneous PACs. These extrastimuli were delivered at a CI of 450 ms from the prior sinus beat and decremented by 50 ms every 10 beats to 200 ms (Supplemental Figure 1B). During PACs at each different CI, echocardiography and LA pressure were recorded simultaneously.

ECHOCARDIOGRAPHY DURING PACED PACs AT DIFFERENT CIs.

During paced PACs, LA mechanics were assessed using high resolution 2-dimensional (2D) strain echocardiography (M5Sc-D probe/Vivid E95 system, GE Healthcare). Echocardiographic images were acquired using a 3.5-MHz transducer placed in the left parasternal area in the right lateral decubitus position. Three-lead electrocardiographic monitoring was applied, and 15 consecutive beats were saved several times per each different PAC CI setting, in cine loop format. Representative cases illustrating the definition of peak LA longitudinal strain, LA dyssynchrony, and LA dyssynergy from 2D strain echocardiography are summarized in Figure 1.¹⁶ Those variables were assessed by 2 experts (S.H., D.B.), and each analysis was performed 3 times and averaged.

FIGURE 1. Representative Cases Illustrating Peak LA Strain, LA Dyssynchrony, and LA Dyssynergy From 2-Dimensional Strain Echocardiography.

Echocardiographic 2-dimensional strain curves including peak left atrial (LA) longitudinal strain, LA dyssynchrony, and LA dyssynergy at a coupling interval (CI) of 250 ms and 400 ms in the modified apical 4-chamber view. Vertical arrows indicate peak LA strain. Peak LA longitudinal strain was defined as average peak strain for 6 segments (white dotted line). Horizontal arrows LA dyssynchrony defined as the time between the earliest and latest peak strain among 6 LA segments. The arrows indicate minimum and maximum strain at global peak LA strain. LA dyssynergy represents the heterogeneity of LA regional myocardial contraction and is defined as maximum regional strain minus minimum regional strain at peak global LA strain. PAC = premature atrial complex.

PEAK LA RESERVOIR STRAIN.

Peak LA reservoir strain was measured using 2D speckle-tracking strain analysis software (EchoPAC version 201; GE Healthcare). A longitudinal strain curve was generated during sinus rhythm and the paced PACs. The strain curve was gated for atrial wall motion among 6 LA segments obtained from the modified apical 4-chamber view at a frame rate of ≥70 frames per second. Peak LA strain was assessed by measuring the average of the first peak longitudinal strain (white dotted line in Figure 1) across 6 LA segments following the paced PACs. The prior QRS onset was set as the reference point.

LA DYSSYNCHRONY.

To quantify mechanical intraatrial dyssynchrony, the difference in the time to peak of the earliest and latest activated segments among the 6 LA segments, following the paced PACs, was measured. Mechanical regional incoordination was assessed during sinus rhythm and at each paced PAC CI.

LA DYSSYNERGY.

LA dyssynergy represents the heterogeneity of LA regional myocardial contraction. At the time of the global peak LA strain, strain values from all 6 LA segments were measured. Dyssynergy was defined as the difference between the maximum strain value and the minimum strain value at the timing of global peak LA strain as defined previously.

LEFT VENTRICULAR DYSSYNCHRONY AND LEFT VENTRICULAR PEAK RADIAL STRAIN.

The radial left ventricular (LV) strain was also assessed using the parasternal short-axis view obtained during sinus rhythm and paced PACs at different CIs. To quantify LV dyssynchrony, the difference in the time to peak of the earliest and latest peak systolic strain, following the paced PACs, among the 6 LV segments was measured. In this context, it was observed that at a shorter CI, the PAC was typically not conducted to the ventricle, resulting in a single hump in the radial strain curve. In contrast, at a longer PAC CI, there was conduction to the ventricle leading to 2 humps in the radial strain curve (Supplemental Figure 2). LV peak radial strain was defined as the mean peak systolic strain in 6 LV segments (Supplemental Figure 2). The endocardial border was traced in end-systole with exclusion of the pericardium, and tracking quality was ascertained visually as well as by the automated algorithm included in the software. In cases of poor tracking, the region of interest was manually readjusted.

PHASE 2: CHRONIC STUDY.

In the phase 2 study, we included 25 swine with a 50% burden of paced PACs distributed in 3 groups: 1) paced PACs at a short CI of 250 ms (Short-PAC group, n = 10); 2) paced PACs at a long CI of 400 ms (Long-PAC group, n = 5); and 3) a control group without pacing (CTRL, n = 10). The chronic study protocol was previously outlined in our recent publication.¹⁵ The long CI of 400 ms was chosen because it allowed continuous PACs at heart rates up to 150 beats/min and was the longest programmable atrioventricular delay (to trigger the paced PAC) on the pacemaker. The short CI of 250 ms was chosen because it was the shortest interval that would allow consistent atrial capture. All animals underwent a comprehensive electrophysiologic and echocardiographic assessment at baseline and after 16 weeks of PAC exposure (terminal study) with the protocol summarized in Figure 2.

FIGURE 2. Study Protocol for Chronic Study.

Baseline and terminal study flowchart describing the protocol for the 3 groups of swine in the study. AF = atrial fibrillation; EP = electrophysiology; ERP = effective refractory period; other abbreviations as in Figure 1.

BASELINE ELECTROPHYSIOLOGY STUDY.

A steerable quadripolar mapping catheter was inserted via a femoral vein sheath. Atrial effective refractory periods (ERPs) were measured at 3 distinct pacing sites: 1) the right atrial (RA) free wall; 2) the right interatrial septum; and 3) the coronary sinus. Single-paced atrial extrastimuli were delivered at a CI of 400 ms from the preceding sensed atrial complex and reduced by 10 ms increments until loss of capture, with a 5-second pause between extrastimuli. The atrial ERP was defined as the longest CI without atrial capture. Following a 10-minute waiting period, AF induction was attempted 3 times from each of the 3 atrial sites (total of 9 attempts) using burst atrial pacing (cycle length 50 ms of 15 seconds). AF sustainability, calculated as the average maximum duration of AF across the 3 sites, and AF inducibility, defined as the proportion of induced AF duration of ≥5 seconds out of the total 9 AF induction attempts, were recorded.

PACEMAKER IMPLANTATION AND PACING PROTOCOL.

An incision was then made on the right neck and blunt dissection used to expose the external jugular vein. Following distal ligation of the vein, an anterior venotomy was performed, allowing for the direct introduction of 2 pacemaker leads. One bipolar active fixation lead (Medtronic 5076) was positioned in the RA appendage for sensing. A second pacing lead was inserted into the distal coronary sinus with the objective of maximizing atrial dyssynchrony (Supplemental Figure 3A), as per our previous study. A biventricular pacemaker (Syncra or Viva CRT-P, Medtronic) was employed to produce paced PACs. The first lead at RA appendage was connected to the atrial port for sensing intrinsic atrial depolarizations, while the second lead at distal coronary sinus was connected to the LV port for sequential atrial pacing and to prevent oversensing. The right ventricular port was plugged.

After a recovery period of 1 week, the pacemakers were programmed from sensing-only (ODO) mode to DDD mode to create paced atrial bigeminy (50% PAC burden). To avoid unnecessary atrial pacing from the sensing lead, the devices were programmed to a low rate (40 ppm). Additionally, the atrial pacing output was set to the minimum level to ensure that even unexpected pacing would not capture the atrial myocardium. The PAC CI (sensed atrioventricular delay on the pacemaker) was set to 250 ms for the short-CI group and to 400 ms for the long-CI group (Supplemental Figure 3B). The pacing output was configured to ensure atrial capture of the bigeminal PACs, while preventing ventricular and phrenic nerve capture. Pacemaker interrogations were conducted monthly to confirm effective sensing, atrial capture, and the absence of AF. Ten swine underwent the same baseline and terminal electrophysiology study and echocardiography protocol and were included as control animals (CTRL group, n = 10).

TRANSTHORACIC ECHOCARDIOGRAPHY.

Transthoracic echocardiography was performed at baseline and terminal stages of the study. The standard settings for transthoracic echocardiography were consistent with those utilized in the acute study. The LA area was measured in a blinded manner using images obtained from a modified apical 4-chamber view at end-systole, with meticulous exclusion of the LA appendage and pulmonary veins. LA function was assessed by quantifying the peak LA reservoir strain as well as by quantifying the LA dyssynchrony using the same 2D speckle-tracking strain analysis method as in the acute study. Additionally, LV function was evaluated by measuring the LV ejection fraction using M-mode in the parasternal short-axis view. Notably, all measurements were made during sinus rhythm.

TERMINAL STUDY.

Following 16 weeks of paced PAC exposure, the electrophysiology study and transthoracic echocardiography were repeated under general anesthesia, adhering to the same protocol as the baseline study. Prior to the study, the pacemakers were deactivated for 24 hours to mitigate any acute effects of frequent PACs. If burst pacing–induced AF did not spontaneously convert to sinus rhythm after 7 minutes, electrical cardioversion was performed. After introducing the deflectable sheath into the LA, the mean LA pressure was obtained during sinus rhythm and coronary sinus pacing at 120/min for 5 swine in the Short-PAC group and 10 swine in the CTR group. The mean aortic pressure was also recorded via the femoral artery sheath during sinus rhythm. Subsequently, a multipolar grid catheter (HD-Grid, Abbott) with 3–3-3-mm interelectrode spacing was introduced into the RA as well as into the LA, and high-density 3-dimensional electroanatomic maps were generated (NavX EnSite Velocity system version 3.0, Abbott). Offline analysis involved evaluating the distribution of the unipolar electrogram voltage across all RA and LA points. Conduction velocity (CV) was determined in the LA anterior and posterior walls using 5 electrogram pairs during coronary sinus pacing at 120 beats/min from each pacing site for each animal. Local activation time was measured perpendicular to the isochrones in areas with minimum isochronal crowding, and the CV was calculated as the surface distance between each point pair divided by the difference in local activation time. The evaluator was blinded to the experimental group.

HISTOLOGICAL ANALYSIS.

Following euthanasia, the swine were dissected, and full-thickness 2-cm² samples were collected from the LA appendage, LA posterior wall, LA anterior wall, and RA appendage. These sections were fixed in buffered formalin and subsequently embedded in paraffin. Staining with Masson’s trichrome was performed on sections from each sample, followed by quantification of fibrous tissue using photomicrographs captured under bright-field microscopy. Fibrosis was quantified by analyzing magnified (×20) images from each section using ImageJ version 1.52 (National Institutes of Health). This analysis involved counting the number of blue-stained pixels, excluding perivascular areas. The fibrous tissue content was expressed as the percentage of field area and averaged across 7 pictures in each site. All analyses were conducted on coded specimens while the evaluator was blinded to the pacing group assignment.

STATISTICAL ANALYSIS.

Continuous variables are presented as mean ± SD for normally distributed data or median (Q1-Q3) for skewed data. Categorical variables were summarized as proportions. To compare means between continuous variables, either an unpaired t test or a 1-way analysis of variance was employed for normally distributed data, while the Kruskal-Wallis test was utilized for skewed data. Normality was assessed using the Shapiro-Wilk test. For analysis of variance, Tukey’s post hoc test was used for multiple comparisons. The statistical analysis was performed using the SPSS for Windows (version 23, IBM). Two-tailed P values <0.05 were considered statistically significant.

RESULTS

PHASE 1: ACUTE STUDY.

A total of 5 swine were included in the acute phase 1 study. All swine successfully underwent simultaneous echocardiography and LA pressure measurements during the paced PACs at different CIs.

ABNORMAL LA MECHANICS ASSOCIATED WITH SHORT VS LONG PAC CIs.

Figure 3 depicts a representative example of LA mechanics and LA pressure during sinus rhythm and at various single-paced PAC CIs. Compared with the sinus beat (Figure 3A), the peak LA strain (Figure 3, top) following the paced PAC beat (orange dotted lines represent the timing of PAC onset on the electrocardiography) was lowest at 250 ms (Figure 3B) and progressively increased with longer CI (Figures 3C and 3D). Mechanical LA dyssynchrony and LA dyssynergy increased with progressively shorter CI. When focusing on simultaneous recordings of LA and aortic pressure (Figure 3, middle), the magnitude of the additional a-wave generated by the PAC, indicated by blue stars in each panel, was substantially smaller during PACs at short CIs (Figure 3B) compared with PACs at long CIs (Figures 3C and 3D). This is likely due to the very limited time allowed for LA filling with short-coupled PACs. However, there was a greater increase in LA pressure from pre-PAC (red dots) to post-PAC (green dots) at shorter CIs (Figure 3B) than was seen at longer CIs (Figure 3D). There was no significant change in LV dyssynchrony during either nonconducted or conducted PACs at all CIs (Figure 3, bottom).

FIGURE 3. Simultaneous 2-Dimensional Strain Echocardiography and LA Pressure During PAC at Different Cis.

The summary of LA mechanics at different PAC CIs is presented at the top. Orange dotted lines represent the timing of PAC onset on the electrocardiogram. In comparison with (A) the sinus beat, (B) the peak LA strain subsequent to the paced PAC beat was diminished during PAC with a short CI of 250 ms but (C, D) was elevated as the CI prolonged. Mechanical LA dyssynchrony and LA dyssynergy were most pronounced during paced PAC with short CIs but demonstrated improvement with longer CI. Examining simultaneous LA pressure (middle), the magnitude of the additional a-wave generated by PAC atrial contraction (blue stars) was substantially smaller during (B) PAC at short CIs compared with (C, D) PACs with long CIs, presumably due to less time for atrial filling. (Middle) LA pressures at the time of a waves before PAC (red dots) and after PAC (green dots). Last, the left ventricular (LV) strain results depicted well-organized LV strain mechanics even across different CIs (bottom of Figure 3). LAP = left atrial pressure; other abbreviations as in Figure 1.

The summary curves in all 5 animals are shown in Figure 4. As the CI of the paced PAC decreases, the average peak LA reservoir strain diminishes (P = 0.003) (Figure 4A) and the LA dyssynchrony increases (P < 0.001) (Figure 4B), leading to less effective atrial relaxation; LA dyssynergy increases with decreasing CI until 350 ms, beyond which it plateaus (P = 0.001) (Figure 4C). Notably, as PAC CIs shorten, there is a significant increase in post-PAC LA pressure (P = 0.001) (Figure 4D). As CIs lengthen, the likelihood of conduction to the ventricle increases, with all swine showing conducted PACs from a CI of 400 ms onward (Supplemental Figure 4). Yet, LV synchrony (P = 0.48) (Figure 4E) and LV peak strain (P = 0.28) (Figure 4F) exhibited no significant changes irrespective of the variations in CIs. In summary, compared with long-coupled PACs, shorter-coupled PACs have worse LA function due to increased atrial dyssynchrony and dyssynergy, leading to greater increases in post-PAC LA pressure. Whether the PACs were conducted or nonconducted had minimal contribution to ventricular function.

FIGURE 4. Summary of Acute LA Mechanics and LV Dyssynchrony at Different PAC Cis.

(A) Peak LA strain, (B) LA dyssynchrony, (C) LA dyssynergy, (D) difference in LA pressures at a-wave between post-PAC and pre-PAC, (E) LV dyssynchrony, and (F) LV peak strain during different PAC CIs. ANOVA = analysis of variance; abbreviations as in Figures 1 and 3.

PHASE 2: CHRONIC STUDY.

Twenty-five swine were included in the chronic phase 2 study. Fifteen swine underwent pacemaker implantation (Short-PAC, n = 10; Long-PAC, n = 5) and successfully established bigeminal atrial pacing while 10 swine served as CTRL group without PACs. As demonstrated in the acute study, PACs were nonconducted to the ventricle in the Short-PAC swine, while PACs were conducted to the ventricle in the Long-PAC swine. The atrial rate was highest in the Short-PAC group, followed by the Long-PAC group, and then the CTRL group (CTRL 88 ± 10 beats/min vs Short-PAC 121 ± 12 beats/min vs Long-PAC 112 ± 7 beats/min; P < 0.001) (Supplemental Figure 5).

CHRONIC ATRIAL STRUCTURAL REMODELING.

The change in the LA area following 16 weeks of paced atrial bigeminy is summarized in Figure 5A. The Short-PAC group exhibited the greatest increase in LA area, followed by the Long-PAC then the nonpaced CTRL group (Δ[terminal baseline]; CTRL 0.9 ± 0.4 cm² vs Short-PAC 5.9 ± 1.2 cm² vs Long-PAC 3.9 ± 0.5 cm²; P < 0.001). Additionally, the decrease in peak LA reservoir strain was most pronounced in the Short-PAC group, followed by the Long-PAC and CTRL groups (Δ[terminal baseline]; CTRL −0.7% ± 4.2% vs Short-PAC −17.3% ± 3.2% vs Long-PAC −12.1% ± 2.1%; P < 0.001) (Figure 5B). Moreover, the Short-PAC group exhibited significant LA dyssynchrony during sinus rhythm compared with the CTRL group (Δ[terminal baseline]; CTRL 3.3 ± 8.5 ms vs Short-PAC 18.3 ± 11.1 ms vs Long-PAC 9.4 ± 2.5 ms; P = 0.005) (Figure 5C). Both the Short-PAC and Long-PAC groups demonstrated no change in LV ejection fraction (Δ[terminal baseline]; CTRL 0.7% ± 3.3% vs Short-PAC −2.5% ± 6.2% vs Long-PAC −1.8% ± 3.9%; P = 0.37) (Supplemental Figure 6).

FIGURE 5. LA Size and Function After Frequent PACs for 16 Weeks.

Change in the (A) LA area, (B) LA peak reservoir strain, and (C) LA dyssynchrony between the baseline and 16-week terminal studies among control (CTRL), Short-PAC, and Long-PAC groups. ANOVA = analysis of variance; other abbreviations as in Figure 1.

CHRONIC ATRIAL ELECTRICAL REMODELING.

After 16 weeks of PACs, the Short-PAC and Long-PAC swine exhibited significantly lower CVs on both the anterior LA wall (CTRL 1.4 ± 0.2 m/s vs Short-PAC 1.0 ± 0.2 m/s vs Long-PAC 1.2 ± 0.2 m/s: P < 0.001) and posterior LA (CTRL 1.4 ± 0.2 m/s vs Short-PAC 1.1 ± 0.1 m/s vs Long-PAC 1.2 ± 0.2 m/s: P < 0.001) compared with CTRL swine (Figures 6A and 6B), with no statistical difference between short-CI and long-CI swine. The change in atrial ERP from baseline to 16 weeks did not differ among the groups (Δ[terminal baseline]; CTRL −3.2 ± 23.7 ms vs Short-PAC −4.0 ± 23.6 ms vs Lat-PAC −8.7 ± 17.9 ms; P = 0.90) (Figure 6C). The distribution of unipolar voltage values in the LA and RA among the entire points is illustrated in Supplemental Figure 7. The acquired points for the use of the analysis did not show differences among the groups (LA: CTRL 2,016 ± 542 vs Short-PAC 2,109 ± 648 vs Long-PAC 1,586 ± 382; P = 0.23; RA: CTRL 2,441 ± 673 vs Short-PAC 2,672 ± 901 vs Long-PAC 1,901 ± 526; P = 0.38). The Short-PAC group had a higher proportion of low-voltage area in both the RA and LA, defined as % area below 1.0 mV, compared with the other groups, indicating a leftward shift in the voltage distribution (LA: CTRL 1.8% ± 1.7% vs Short-PAC 17.2% ± 11.8% vs Long-PAC 7.8% ± 5.7%; P = 0.004; RA: CTRL 2.7 ± 0.9% vs Short-PAC 15.1% ± 8.6% Long-PAC 10.2% ± 5.8%; P = 0.02).

FIGURE 6. LA Electrophysiology After Frequent PACs for 16 Weeks.

(A) LA anterior wall and (B) LA posterior wall conduction velocity (CV) during the terminal study in each group. CV was measured during atrial pacing from the coronary sinus at 120 beats/min. (C) Change in the atrial effective refractory period (ERP) between the baseline and 16-week terminal studies in each group. The atrial ERPs are obtained by the average of the ERPs on the right atrial free wall, right atrial mid-septum, and coronary sinus. *P < 0.05; **P < 0.001. Abbreviations as in Figures 1 and 5.

CHRONIC CHANGE IN HEMODYNAMICS.

At the terminal study, the Short-PAC group exhibited a significantly higher mean LA pressure than the CTRL group, both during sinus rhythm and during pacing at the same atrial rate of 120 beats/min from coronary sinus (Supplemental Figures 8A and 8B), with no significant change in mean aortic pressure (Supplemental Figure 8C).

AF INDUCTION.

After 16 weeks of PACs, the average duration of induced AF was increased in both the Short-PAC and Long-PAC groups compared with the CTRL group (Δ[terminal baseline]; CTRL −1 second [Q1-Q3: −16 to 7 seconds] vs Short-PAC 131 seconds [Q1-Q3: 30–192 seconds] vs Long-PAC 30 seconds [Q1-Q3: 11–93 seconds]; P < 0.001) (Figure 7A). Furthermore, the increase in the AF inducibility between terminal and baseline study was greatest in the Short-PAC group (Δ[terminal baseline]; CTRL −2.2% ± 16.2% vs Short-PAC 49.3% ± 13.0% vs Long-PAC 29.0% ± 6.4%; P < 0.001) (Figure 7B). All induced AF events comprised irregular atrial activity, with no regular atrial tachyarrhythmias observed.

FIGURE 7. AF Maintenance After Frequent PACs for 16 Weeks.

(A) Atrial fibrillation (AF) sustainability: the change in duration of induced AF between baseline and 16-week terminal studies in each group. AF sustainability is defined as the average of maximum AF duration obtained on the right atrial free wall, right atrial mid-septum, and coronary sinus. Cardioversion was performed after 7 minutes (420 seconds) of sustained AF. (B) AF inducibility: the change in the percentage of inducible AF duration of ≥5 seconds among the 9 total AF induction attempts between the baseline and 16-week terminal studies in each group. Abbreviations as in Figures 1 and 5.

HISTOPATHOLOGY.

The histological fibrosis in the LA and RA in each group is summarized in Figure 8. The Short-PAC group exhibited the highest degree of fibrosis, followed by the Long-PAC group and then the CTRL group. Specifically, the Short-PAC group showed significantly higher % fibrosis than the CTRL group in all locations. The Long-PAC group demonstrated significant higher fibrosis than the CTRL group on the LA posterior wall (CTRL 4.0% ± 1.5% vs Short-PAC 8.6% ± 1.0% vs Long-PAC 6.8% ± 1.0%; P < 0.001) (Figure 8E) and the LA anterior wall (CTRL 3.9% ± 1.7% vs Short-PAC 7.7% ± 1.6% vs Long-PAC 6.4% ± 1.5%; P < 0.001) (Figure 8F).

FIGURE 8. Histopathology Analysis After Frequent PACs for 16 Weeks.

Representative histological slides of the LA posterior wall for the (A) control (CTRL), (B) Short-PAC, and (C) Long-PAC groups. The numbers indicate the mean percentage of fibrosis in the LA posterior wall for each group. Differences in the histological % fibrosis in the (D) LA appendage, (E) LA posterior wall, (F) LA anterior wall, and (F) right atrial (RA) appendages by group. *P < 0.05; **P < 0.001. Abbreviations as in Figures 1 and 5.

DISCUSSION

In our swine model of paced PACs, we found that: 1) PACs with progressively shorter CIs resulted in lower peak LA strain, greater LA dyssynchrony, and greater LA dyssynergy; and 2) prolonged exposure to frequent short-CI PACs over 16 weeks led to a greater degree of structural fibrotic remodeling and induced AF compared with long-CI PACs (Central Illustration). This suggests that the timing of PACs plays an important role in atrial mechanical function, with shorter CIs potentially causing more pronounced electrical and structural remodeling.

Prior epidemiology studies in humans have found a correlation between frequent PACs and the subsequent development of AF^1–8; however, whether the PACs were causative, or an epiphenomenon, was not known. In addition, it was not clear whether all frequent PACs are similarly profibrillatory or whether certain PACs are more profibrillatory than others. Our prior work demonstrated that those PACs originating from the lateral LA leading to greater LA dyssynchrony were more likely to lead to atrial remodeling and AF.¹⁵ In the current study, we demonstrate that PACs with shorter CI are more likely to lead to profibrillatory atrial remodeling.

Our study does not directly address the question of the basic process whereby frequent PACs lead to atrial remodeling and AF but can form hypotheses from these data. The underlying driver of atrial remodeling seems to be dyssynchronous atrial myocardial mechanics, resulting in reduced peak LA reservoir strain, particularly as the CI shortens. Our model suggests that short-CI PACs contribute to heterogeneity in relaxation and a reduction in total reservoir function. This reduction likely leads to elevated mean LA pressure, potentially due to increased LA wall stretch—paralleling the effects seen with frequent premature ventricular contractions contributing to ventricular cardiomyopathy, as demonstrated in both preclinical and clinical studies.^17–19 We hypothesize that this chronically elevated LA pressure may lead to progressive LA enlargement and fibrotic remodeling that supports the maintenance of AF. Indeed, the short-CI PAC group exhibited elevated LA pressure during sinus rhythm in the terminal study, suggesting presence of a chronical pressure overload. Recent insights into atrial remodeling mechanisms in response to pressure overload emphasize the interconnected nature of electrical and structural changes in the LA.²⁰ Chronic pressure overload can lead to down-regulation of rapid conduction genes and upregulation of profibrotic genes that contribute to atrial electrical and structural remodeling. One might argue that short-coupled PACs are more likely to lead to elevated atrial pressure and remodeling than long-CI PACs because the short-CI PACs are nonconducted and may lead to atrial contraction against a closed mitral valve. However, we have demonstrated that the short-CI PACs actually generate very little pressure, likely because: 1) the PACs occur so early that little LA filling has occurred; and 2) the sarcoplasmic reticulum may not have had time to fully resorb calcium before depolarization occurs. The increase in atrial dyssynchrony observed with short-CI PACs may be related to heterogeneous states of calcium resorption into the sarcoplasmic reticulum throughout the atrium, leading to uncoordinated and ineffective mechanics. Our findings of increased dyssynchrony and fibrosis with short CI are similar to what we found in a swine model of paced premature ventricular contractions.²¹ It is likely that dyssynchrony of any cardiac chamber may lead to ensuing structural myopathy. Heretofore, this has been under-recognized in the atrium because the ensuing atrial myopathy less readily leads to symptoms. We have previously shown, in a patient with frequent short-coupled PACs, that elimination of the PACs through catheter ablation led to a significant improvement in atrial mechanical function.¹¹

CLINICAL IMPLICATIONS.

These insights may have important implications for the management and prevention of atrial myopathy and AF in clinical practice. Upstream therapy to prevent AF has had limited success.^22–24 Eliminating frequent PACs may be a novel therapy to prevent future AF, and a new diagnosis of “pre-AF” has been established in the recent American Heart Association AF management guidelines.²⁵ Consistent PACs from the lateral LA with a shorter CI may be most likely to lead to atrial remodeling and could be detected with ambulatory monitoring. Potential approaches to prevent atrial remodeling may include pharmacological interventions, lifestyle modifications, or potentially invasive procedures such as catheter ablation. Further validation of these findings in humans is necessary; closer monitoring of patients with frequent PACs may be prudent.

STUDY LIMITATIONS.

While the use of a swine model may not perfectly replicate human physiology, it allowed us to prospectively assess the direct effects of high-burden PACs over a 16-week period under controlled conditions—an approach that would be ethically challenging in humans. Although a 50% PAC burden is uncommon clinically, this model enabled us to investigate whether PACs could induce atrial remodeling over a realistic time frame. Further research is needed to determine the PAC burden and duration required in humans to develop atrial myopathy. Additionally, the findings may not be generalizable to all patient populations, and caution should be exercised when applying the results to diverse clinical scenarios. Further research is warranted to elucidate the precise mechanisms underlying the differential effects of short-CI PACs vs long-CI PACs on atrial remodeling and AF development. Deeper mechanistic insights may guide the development of novel therapeutic approaches targeting PAC-mediated arrhythmogenesis in clinical practice.

CONCLUSIONS

Our findings from this swine model demonstrate a direct association between shortening of PAC CI and acute atrial mechanical dysfunction and dyssynchrony. Chronic exposure to short-coupled PACs resulted in more pronounced atrial fibrosis and more prolonged induced AF. These observations suggest that frequent, short-coupled PACs pose the highest risk for developing LA myopathy and AF. These insights underscore the importance of understanding the impact of PAC characteristics on atrial remodeling and arrhythmogenesis.

PERSPECTIVES.

COMPETENCY IN MEDICAL KNOWLEDGE:

Findings from this swine model demonstrate a direct association between shortening of the PAC CI and acute atrial mechanical dysfunction and dyssynchrony. Moreover, chronic short-CI PACs resulted in more pronounced atrial fibrosis and more prolonged induced AF. These observations suggest that frequent, short-coupled PACs pose the highest risk for developing LA myopathy and AF.

TRANSLATIONAL OUTLOOK:

Eliminating frequent PACs may be a novel therapy to prevent future AF and the recent American Heart Association guidelines have established a new diagnosis of “pre-AF.” Consistent PACs from the lateral LA with a shorter CI may be most likely to lead to atrial remodeling and could be detected with ambulatory monitoring. Further validation of these findings in humans is necessary; close monitoring of patients with frequent PACs may be prudent.

ABBREVIATIONS AND ACRONYMS

AF

atrial fibrillation

CI

coupling interval

CV

conduction velocity

ERP

effective refractory period

LA

left atrium/atrial

LV

left ventricle/ventricular

PAC

premature atrial complex

RA

right atrium/atrial

2D

2-dimensional