Functional and Structural Remodeling as Atrial Fibrillation Progresses in a Persistent Atrial Fibrillation Canine Model

Abstract

BACKGROUND

Contractile, electrical, and structural remodeling has been associated with atrial fibrillation (AF), but the progression of functional and structural changes as AF sustains has not been previously evaluated serially.

OBJECTIVES

Using a rapid-paced persistent AF canine model, the authors aimed to evaluate the structural and functional changes serially as AF progresses.

METHODS

Serial electrophysiological studies in a chronic rapid-paced canine model (n = 19) prior to AF sustaining and repeated at 1, 3, and 6 months of sustained AF were conducted to measure changes in atrial conduction speed and direction. Cardiac late gadolinium enhancement magnetic resonance imaging was performed prior to and following sustained AF to evaluate structural remodeling.

RESULTS

As AF progressed, the overall area of the left atrium with fibrosis increased. Over time, conduction speeds slowed, with speeds decreasing by 0.15 m/s after 3 months and 0.26 m/s after 6 months of sustained AF. Regions that developed fibrosis experienced greater slowing compared with healthy regions (0.32 ± 0.01 m/s decrease vs 0.21 ± 0.01 m/s decrease; P < 0.001). Conduction directions became more aligned (conduction direction heterogeneity decreased from 19.7 ± 0.1° to 17.5 ± 0.1° after 6 months of sustained AF; P < 0.001). Fibrotic regions had a greater decrease in conduction direction heterogeneity (2.7 ± 0.3° vs 2.0 ± 0.2°; P = 0.008).

CONCLUSIONS

As AF progressed, functional changes occurred globally throughout the left atrium. Conduction speed slowed, and conduction directions became more aligned over time, with the greatest changes occurring within regions that developed fibrosis.

Atrial fibrillation (AF) is the most prevalent and common cardiac arrhythmia worldwide.^1–3 AF is a conduction abnormality that causes the atria to conduct without a normal rhythm, and the arrhythmia is linked to electrical, structural, and contractile remodeling.^4–7 Because of this remodeling, AF burden worsens over time.^8,9 There is a great need to understand how the arrhythmia develops and sustains so that better treatment and ablation strategies may be developed.

The functional and structural changes in the atria as AF progresses remain poorly understood. To some degree, prior work has linked functional characteristics such as conduction slowing and structural remodeling such as fibrosis with progression of AF.^10,11 However, these studies compared structural and functional differences between AF-induced animals compared with healthy control animals, lacking serial measurements to examine temporal changes. Previous studies have shown, both clinically and in animal models, that structural remodeling may be linked to slower conduction.^11–16 These studies were performed at single variable time points following AF onset but did not comprehensively evaluate conduction changes as AF progresses. A significant contribution of our study is the evaluation of how conduction speeds change serially as AF progresses and structural remodeling occurs.

In addition to conduction slowing, AF is associated with changes in conduction anisotropy and changes in conduction dynamics.^11,15–18 This may be the result of underlying structural remodeling or due to the natural time course of the disease.^10,11,16 As AF progresses, there is increased atrial fibrosis due to structural remodeling.^9,10,19 It is expected that conduction slowing and changes in conduction patterns may have more drastic effects in regions that developed fibrosis compared with healthy regions of the atria.^11,16

Performing serial follow-up electrophysiological (EP) studies as AF progresses, we aimed to characterize the conduction changes in the left atrium (LA) and relate these functional changes with underlying structural remodeling. Conduction direction, speeds, and dynamics were evaluated over time, with emphasis on the differences in changes between regions that remained healthy and regions that developed fibrosis. The aim of the study was to provide insights on the functional changes in the atria attributable to AF and to relate these to structural remodeling. Regions that developed fibrosis were expected to demonstrate impaired conduction compared with regions that remained healthy.

METHODS

All animal studies were performed in accordance with the Guide for the Care and Use of Laboratory Animals,²⁰ with a protocol approved by the Institutional Animal Care and Use Committee at the University of Utah.

ANIMAL MODEL.

Purpose-bred mongrel hounds (1–2 years of age, weight 25–35 kg; Oak Hill Genetics) underwent rapid atrial pacing to induce AF.¹⁰ For pacemaker implants, EP studies, and magnetic resonance imaging (MRI) scans, animals were anesthetized with propofol (5–8 mg/kg) and then intubated for general anesthesia with isoflurane (1.5%-4%). Throughout the procedures, animals were mechanically ventilated.

PACING PROTOCOL.

To induce AF, a screw-in pacing lead (model 4076, Medtronic) was implanted into the high right atrium and attached to a neurostimulator (InterStim 3023, Medtronic) to rapidly pace the atria. The pacing protocol was set to pace the atria at 50 Hz at 2 times the diastolic pacing threshold, alternating between 2 seconds pacing and 2 seconds off pacing. Following pacing, electrocardiographic checks were performed weekly to evaluate whether AF was sustained. Electrocardiographic checks were done 20 minutes after the pacing protocol was turned off. Once AF was observed to sustain after 20 minutes of the pacemaker turned off (2–8 weeks after initial pacing stimulus), the AF timeline was started for the animal. At this point, the pacing protocol was adjusted to pace for 2 seconds, followed by 1 minute off. The adjusted pacing protocol allowed AF to progress without continuous pacing, but the extra stimulus ensured that the animal returned to AF in the event of spontaneous sinus rhythm. Rhythm checks were performed weekly to confirm that the animals remained in AF. Animals were determined to have sustained AF if they did not spontaneously revert back to normal sinus rhythm during the 20-minute blanking period with the pacemaker turned off at the weekly rhythm checks.

EP STUDIES.

EP studies were performed prior to the initiation of the pacing protocol and repeated serially after AF was sustained. All animals underwent EP procedures at 3 and 6 months of sustained AF. A subset of animals also underwent EP procedures at 1 month of sustained AF to evaluate early functional changes as AF progresses.

Animals were anesthetized as previously described, and the pacemaker was turned off. Femoral arterial access was obtained to monitor blood pressure. Femoral venous access was obtained for catheter insertion. A transeptal puncture was made using a transeptal puncture needle (HeartSpan, Merit Medical) to access the LA under fluoroscopy and intracardiac echocardiographic guidance (Acuson AcuNav intracardiac echocardiographic catheter, Siemens Healthineers). LA electrical mapping was performed using the Rhythmia Cardiac Mapping System (Boston Scientific). A decapolar catheter (Webster Duo-Decapolar Catheter, Biosense Webster) was placed in the right atrium to serve as the stable reference signal. A high-density, 64-electrode basket Orion catheter (Boston Scientific) was used to map the LA in normal sinus rhythm and in paced rhythms. The animals were cardioverted to normal sinus rhythm prior to mapping. Activation times were annotated using bipolar electrograms (EGMs). The local activation times for the sinus and paced maps were measured as the timing difference between the activations on the roving Orion catheter and the reference catheter.

To evaluate conduction dynamics, pacing was initiated through the implanted pacemaker lead to ensure that pacing origin remained the stay same between the serial studies. The pacemaker was programmed to pace constantly at different cycle lengths (384, 285, 256, and 194 ms, determined by the allowable pacing frequencies on the neurostimulator). Local activation time maps for each pacing cycle length were mapped with the bipolar EGM signals.

For effective refractory period (ERP) testing, an S1-S2 protocol was used. S1 was set at 400 ms and S2 decremented from 200 to 50 ms in 10-ms intervals until the ERP was reached or AF was stimulated.

MRI.

A late gadolinium enhancement (LGE) MRI scan was performed to evaluate the tissue characteristics in the atria, as previously described.^10,21 The scans were carried out on a 3-T MRI scanner (Siemens Healthineers) with a 3-dimensional inversion recovery, respiration-navigated, electrocardiographically gated, gradient-echo pulse sequence. Images were acquired about 15 minutes following the injection of the contrast agent gadolinium (0.1 mmol/kg Prohance, Bracco). The image acquisition parameters were a voxel size of 1.25 × 1.25 × 2.5 mm (reconstructed to 0.625 × 0.625 × 1.25 mm), repetition time of 3.1 ms, echo time of 1.4 ms, and flip angle of 13° The inversion time (270–330 ms) for each scan was determined using a scout sequence prior to the LGE MRI scan to null the myocardium.

The LGE MRI scan was semiautomatically segmented using CorView (University of Utah) to label the LA wall tissue.^22,23 The LA wall intensity values were scaled to SDs away from the mean of the lower 50% of all wall intensity values to report an MRI score. An MRI score <1 (any MRI intensity value <1 SD above the mean of the lower 50% of all wall intensities) was labeled as healthy tissue. MRI scores >3 were labeled as fibrotic region, with the border zone region labeled as areas with MRI scores of 1 to 3. A 3-dimensional volume of the LA geometry was created from the LGE MRI scan.

Each animal underwent an LGE MRI scan after 6 months of sustained AF, and a subset (n = 18) underwent LGE MRI scans prior to initiating the AF pacing protocol. The 6-month follow-up scan was performed within a week of the 6-month EP procedure. The baseline and 6-month MRI scans were compared, and the fibrotic percentage in the LA body (excluding the pulmonary veins and the appendage) was measured and compared. The fibrotic percentage in differences from baseline to 6 months of sustained AF in different regions of the LA (posterior, anterior, lateral, septal, and roof) were also compared.

CONDUCTION VELOCITY.

The activation times were projected onto a 3-dimensional LGE MRI volume for calculating the conduction velocity. Conduction velocity was calculated using an inverse-gradient, wave front propagation method.^24,25 For each point in the LA, a plane was estimated using the nearby points within 3 mm. With the activation times from all the nearby points projected onto the estimated plane, a conduction wave front was estimated. The resulting conduction speed and direction was assigned to each point on the LA geometry. To evaluate conduction propagation, a conduction direction heterogeneity measurement was performed. For each point, the conduction directions of the neighboring region within 3 mm were compared, with the conduction direction heterogeneity being the average angle difference measured between the point and its neighboring conduction directions. A heterogeneity of conduction index was calculated by determining the timing difference between each point from its nearest neighbors.^26,27 The index provides a measure of the amount of slow conduction patches. A higher index value would indicate more slow conduction patches. The conduction measurements between studies were compared.

HISTOLOGY.

Tissue samples of the LA were taken after fixation for Masson’s trichrome staining. The samples are prepared as previously described.¹¹ Briefly, tissue samples in high- and low-intensity regions labeled by MRI were taken. The tissue samples were paraffin embedded and sliced with 4-μm slice thickness, beginning 500 μm below the epicardium. The histology slides were processed to calculate fibrosis percentage using a custom MATLAB (The MathWorks) script.

STATISTICAL ANALYSIS.

Conduction speed and direction heterogeneity were compared in the healthy, border zone, and fibrotic regions for each study time point and across the different studies. Data are reported as mean ± SE. Correlations between continuous variables were assessed using a linear mixed model to account for variation within animals. One-way analysis of variance with fixed effects was used to assess differences in conduction between studies and between underlying tissue characteristics. The Kolmogorov-Smirnov test was used to determine if the distributions of 2 samples were significantly different, with P values <0.001 considered to indicate statistical significance.

RESULTS

A total of 19 animals (10 males, 9 females) were used in this study. All 19 animals underwent EP studies at baseline before AF was induced, after 3 months of sustained AF, and after 6 months of sustained AF. A subset of 7 animals underwent EP studies after 1 month of sustained AF. All animals underwent LGE MRI within 1 week of the 6-month EP study, and a subset of animals (n = 18) underwent LGE MRI scans prior to AF induction. In total, 184 maps were assessed, with a median of 6,390 EGMs per map (Q1-Q3: 4,627-8,358 EGMs per map).

CONDUCTION VELOCITY AND FIBROSIS.

After 6 months of AF, fibrotic regions had slower conduction and lower conduction direction heterogeneity. Figure 1 illustrates an example of a single animal. Overall conduction speed was 0.75 ± 0.01 m/s, with conduction speeds significantly slower in regions that had structural remodeling compared with healthy regions in the atria (0.78 ± 0.01 m/s in the healthy region, 0.75 ± 0.01 m/s in the border zone, 0.68 ± 0.01 m/s in the fibrotic region; P < 0.001 among all regions) (Figure 2A). There was a negative correlation (coefficient = −0.020; 95% CI: −0.023 to −0.017; P < 0.001) between MRI score and conduction speed. Conduction direction heterogeneity overall in the LA after 6 months of AF was 17.5 ± 0.1°. Regions with increased structural remodeling exhibited less conduction direction heterogeneity compared with healthy regions (17.7 ± 0.2° in the healthy region, 17.7 ± 0.1° in the border zone, and 16.5 ± 0.2° in the fibrotic region; P < 0.001 among healthy and fibrotic regions and between border zone and fibrotic) (Figure 2B). Conduction direction heterogeneity had a negative correlation (coefficient = −0.201; 95% CI: −0.308 to −0.094; P < 0.001) with MRI score. The Kolmogorov-Smirnov test to evaluate differences in distribution showed statistical significance of the conduction speed distribution between the fibrotic and healthy regions and the fibrotic and border zone regions (P < 0.001). The conduction direction heterogeneity distribution was statistically significant between the healthy and fibrotic region and the border zone and fibrotic regions (P < 0.001) but not between the healthy and border zone regions (P = 0.008) (Supplemental Figure 1).

FIGURE 1. Conduction and Structural Remodeling After 6 Months of Atrial Fibrillation: Example of 1 Animal.

(A) Fibrosis map from late gadolinium enhancement magnetic resonance imaging (MRI) with regions of high intensity (score of 1, low intensity; score of 3, high intensity) corresponding with regions of (B) slower conduction speed and (C) lower conduction direction heterogeneity.

FIGURE 2. Conduction Speed and Direction Heterogeneity After 6 Months of Sustained Atrial Fibrillation.

(A) Conduction speeds are slower in regions that developed fibrosis. (B) Conduction directions are more aligned within the fibrotic regions. *P < 0.001.

CONDUCTION CHANGES AS AF PROGRESSES.

As AF progresses, conduction speed decreases (Figure 3), and conduction directions become more aligned. Beginning at 1 month after sustained AF, there were markedly slower conduction speeds throughout the LA. The overall speed in the LA decreased by 0.15 m/s. After 3 and 6 months of AF, conduction speed remained slow compared with baseline. There was no overall conduction slowing between 1 and 3 months of AF, but as AF progressed beyond 3 months toward the 6-month time point, there was a decrease in speed of 0.11 m/s (decreases of 0.15 m/s at 3 months and 0.26 m/s at 6 months) (Figure 4A). Conduction slowing occurred diffusely throughout the atria. However, conduction speeds slowed down more in regions that developed fibrosis compared with regions that remained healthy (speeds decreased by 0.32 ± 0.01 m/s in fibrotic regions compared with 0.21 ± 0.01 m/s; P = 0.001) (Figure 4B). The conduction speed distributions were statistically significantly different between the baseline to 1-month follow-up study (P < 0.001) and between the 3- and 6-month follow-up studies (P < 0.001) but not between the 1- and 3-month follow-up studies (P = 0.063) (Supplemental Figure 2).

FIGURE 3. Conduction Speeds Over Time Decrease as Atrial Fibrillation Progresses.

Posterior view of conduction speeds at baseline, 1 month, 3 months, and 6 months of atrial fibrillation from 1 animal.

FIGURE 4. Functional Changes Over Time as AF Progresses.

(A) Average conduction speed decreases after 1 month of sustained atrial fibrillation (AF) and continue to decrease as AF progresses. (B) Speed decrease within the fibrotic region is greater than the healthy regions. (C) Conduction direction heterogeneity lowers as propagation directions become more aligned over time. (D) The conduction directions are more aligned over time within the fibrotic region. (E) Heterogeneity of conduction index, evaluating the distribution of conduction speeds in a local area, increases over time, suggesting that there are more slower conduction regions as AF progresses. (F) The heterogeneity of conduction index increases over time in all regions, suggesting that slow conducting zones exist diffusely throughout the left atrium. *P < 0.001 and +P < 0.001 between healthy and fibrotic regions.

Overall conduction direction heterogeneity became more aligned gradually over time, going from 19.7 ± 0.1° prior to AF initiating to 17.5 ± 0.1° after 6 months of sustained AF (P < 0.001) (Figure 4C). The decrease in conduction direction heterogeneity is greater in the regions that developed fibrosis compared with the healthy regions (decrease of 2.0 ± 0.2° in the healthy region and decrease of 2.7 ± 0.3° in the fibrotic regions; P = 0.008) (Figure 4D).

Evaluating the heterogeneity of conduction index, a score measuring the local differences in conduction propagation speed, there was an increase in index over time as AF progressed (Figure 4E). There was no statistical difference in the heterogeneity of conduction index between healthy and fibrotic regions after 6 months of sustained AF (Figure 4F). There was heterogeneity in the conduction propagation speeds in the healthy and fibrotic regions, with greater heterogeneity of conduction over time. However, though the heterogeneity of conduction was the same between the healthy and fibrotic regions, the overall conduction speeds were slower in fibrotic regions.

Functional changes did not occur specifically to certain regions, with conduction slowing occurring and more aligned conduction directions observed throughout the LA. After 6 months of sustained AF, conduction slowing occurred throughout the atria (Figure 5). The change was statistically significant (P < 0.01) for all regions of the LA except for the anterior region. Conduction directions were also more aligned in all regions except for the septal wall, where there was a slight increase in conduction direction heterogeneity (Figure 5).

FIGURE 5. Functional Changes in Different Regions of the Atria.

(A) Conduction speed changes from baseline to after 6 months of sustained atrial fibrillation. Speeds slow throughout the atria. (B) Conduction direction heterogeneity becomes more aligned throughout the atria except for the septal region. *P < 0.001.

The ERP at the baseline recording was measured to be 70 to 110 ms throughout the LA for all animals. After 1 month of sustained AF, for all animals, the ERP was <50 ms or AF was induced prior to reaching the ERP. The lowering of the ERP indicates that electrical remodeling had set into the LA within the 1 month of sustained AF.

CONDUCTION DYNAMICS.

Over time there were conduction restitution curve differences (Figure 6). The restitution curves from the baseline and 6-month studies illustrate differences in the conduction dynamics over time as AF progressed. The slope of the restitution curve (between the cycle lengths of 384 and 194 ms) dampened at the 6-month time point (slopes of 1.3 m/s² at baseline and 0.7 m/s² at 6 months). Breaking down the restitution curve to show the conduction dynamics in regions that developed fibrosis and regions that did not, the restitution curves for the different underlying tissue characteristics were different. Compared with the baseline time point, the speeds for all pacing cycles slowed after 6 months. The slope of the curve in the fibrotic regions was dampened compared with the healthy region (0.8 m/s² in the healthy region and 0.6 m/s² in the fibrotic region). The distributions of the conduction speeds at different paced cycle lengths became closer over time (Supplemental Figure 3).

FIGURE 6. Conduction Restitution Curve Changes Over Time as AF Progresses.

(A) Overall restitution curve over time, with the slope decreasing after atrial fibrillation (AF) sustains. Restitution curves within the healthy, border zone, and fibrotic regions at (B) baseline, (C) 3 months of sustained AF, and (D) 6 months of sustained AF. (E) Posterior view example of conduction speeds at different paced cycle lengths over time for 1 animal.

STRUCTURAL REMODELING.

The 6-month post-AF LGE MRI studies showed significantly more fibrotic area in the LA than the baseline MRI (8.98% ± 7.07% to 21.8% ± 12.9%; P < 0.001) (Figure 7). The increase in fibrosis occurred throughout the LA and was not isolated to specific regions (Figure 7D). Masson’s trichrome staining of histology samples taken from high-intensity regions identified by LGE MRI exhibited larger fibrosis content than samples taken in healthy regions. Regions labeled as fibrotic by LGE MRI showed significantly more histologic fibrosis than low-intensity regions (6.41% ± 3.97% vs 3.68% ± 1.56%; P < 0.001) (Figure 8).

FIGURE 7. Structural Remodeling Develops Over Time.

Example from 1 animal of the posterior view of the fibrosis map on the 3-dimensional late gadolinium enhancement (LGE) volume acquired from LGE magnetic resonance imaging (MRI) (A) prior to atrial fibrillation (AF) initiation and (B) after 6 months of sustained AF. (C) Area of the left atrium (LA) with fibrosis as determined by LGE MRI increased after 6 months of sustained AF. (D) Percentage labeled fibrotic in different regions of the LA. There is an increase in fibrosis throughout the LA. #P = 0.07, +P = 0.11, and *P < 0.001.

FIGURE 8. Masson’s Trichrome Stain of Histology Samples Taken After 6 Months of AF.

(A) Fibrosis percentage in histology samples taken from high-intensity (fibrotic) and low-intensity (healthy) regions of late gadolinium enhancement (LGE) magnetic resonance imaging (MRI). Fibrotic regions had higher fibrosis content on histology samples. (B) Fibrosis percentage in histology samples from different regions of the atria. (C) Example of tissue samples taken from 1 animal in fibrotic and healthy regions as identified by LGE MRI. There was increased fibrotic content (blue stain) in samples taken from regions that have been labeled fibrotic (MRI score >3) compared with healthy regions (MRI score <1). *P < 0.001. AF = atrial fibrillation.

DISCUSSION

Atrial remodeling is known to be a key contributor to sustaining AF. Our study is the first to examine functional remodeling as AF progresses in a longitudinal study and relate it to structural remodeling. The key findings of this study illustrated the diffuse functional changes in the LA as soon as 1 month after sustained AF. Conduction slowing occurred throughout the LA as AF progressed, but the regions that developed fibrosis had the greatest conduction slowing (Central Illustration). Conduction direction heterogeneity was observed to change over time, with the fibrotic regions developing more aligned conduction directions. The conduction slowing and increased unidirectional conduction propagation occurring diffusely throughout the atria, but more so in the fibrotic regions, may allow AF to sustain.

CENTRAL ILLUSTRATION. Functional and Structural Remodeling as AF Progresses.

Conduction speeds at (A) baseline prior to atrial fibrillation (AF) induction and (B) after 6 months of sustained AF. (C) The fibrosis map from late gadolinium enhancement magnetic resonance imaging (MRI). An MRI score (scaled MRI intensity) of >3 is labeled fibrosis. (D) Decrease of conduction speed after 6 months of sustained AF. (E) Slowing occurs within the fibrotic region but also throughout the atria. *P < 0.001.

CONDUCTION VELOCITY AND FIBROSIS.

In agreement with previous studies, conduction speeds were slowed in regions that have structural remodeling as revealed with LGE MRI.¹⁴ Slowed conduction speeds have been associated with AF. Slowing of conduction may allow re-entrant circuits to develop as the conduction wave front may now activate repolarized regions. Fibrotic regions may provide sufficient impediment to cardiac tissue fiber connections, leading to slowed conduction.

Previous reports of explanted hearts have indicated that there is greater anisotropy in AF animals and that these anisotropic regions may allow AF to sustain, though these studies did not take into account the underlying fiber orientation when measuring anisotropy.^11,17,18 Instead, the direction of fastest conduction was labeled as the longitudinal direction. The anisotropy measurement made refers to the alignment of conduction direction rather than the preferential flow along a particular fiber direction. To describe the same phenomenon, we used a measure of the conduction direction heterogeneity. This was possible with the high-density mapping done in the entire LA chamber. Examining the difference of angles between conduction directions provides an indication of how directional the wave front propagation is. Regions of structural remodeling exhibited lower conduction direction heterogeneity compared with healthy tissue regions, indicating that conduction propagation was more aligned in the regions of structural remodeling. Conduction direction heterogeneity was inversely related to previous measures of anisotropy, meaning that lower conduction direction heterogeneity scores correspond with previous higher anisotropy measurements. Agreeing with previous findings, we identified that regions of fibrosis had more aligned conduction directions.

In summary, after 6 months of sustained AF, structurally remodeled regions had slower conduction with increased unidirectional conduction propagation. Slower conduction speeds may allow propagation to meander through and activate previously repolarized tissue, while aligned conduction direction may allow unidirectional block to occur. Coupled together, slowed conduction and unidirectional block may allow a re-entrant circuit to develop.²⁸ The functional changes in the structurally remodeled regions indicate that fibrotic regions may provide a substrate for AF to sustain.

CONDUCTION CHANGES AS AF PROGRESSES.

A significant contribution of this study were the longitudinal measurements relating the functional changes to the underlying structural remodeling. Our study revealed that slowed conduction manifested early on, after 1 month of AF, with conduction slowing throughout the atria. The greatest conduction slowing during this period was found in regions that developed fibrosis (Figure 4). Early conduction slowing may be attributed changes in connexin-43 expression and initial increases in atrial fibrosis developing in these regions.²⁹

From 1 to 3 months of sustained AF, there was slight continued conduction slowing, with another drop in conduction speeds occurring between the 3- and 6-month studies. Conduction slowing during the 3- to 6-month period again manifested throughout the LA, illustrating the diffuse functional remodeling occurs throughout the LA. Regions of fibrosis did experience the greatest conduction slowing as AF progressed.

Not only was there conduction slowing over time, but there was also increased alignment in conduction direction over time as AF progressed, indicated by the conduction direction heterogeneity decrease. Conduction propagation became more aligned throughout the atria, suggesting that functional changes were occurring globally. However, the fibrotic regions experienced the greatest decrease in conduction direction heterogeneity, with a greater increase in alignment of conduction directions over time compared with the healthy regions.

Conduction direction alignment may manifest because of interstitial fibrosis developing and interrupting the transverse connections of the tissue. Previous reports have shown that transverse speeds are slower in regions of atrial fibrosis, leading to increased unidirectional conduction propagation.^11,17,30 The fibrotic regions may provide a substrate for increased unidirectional propagation, leading to increased potential for re-entrant circuit developing and sustaining AF.

With the heterogeneity of conduction index, slow conduction regions were evaluated. There was an increase as AF progressed, suggesting that there were more slower conduction regions over time. The heterogeneity of conduction index increased in the healthy and the fibrotic regions, so slower conduction was manifesting throughout the atria and was not solely confined to the fibrotic regions. This result is consistent with previous works suggesting that as AF sustains, there is an increase in conduction abnormalities.^26,27

As AF progressed, functional changes in the atria occurred globally and continued to change over time, with continued conduction slowing and increased conduction direction alignment. The greatest functional changes were contained in the regions that developed fibrosis. The longitudinal study confirmed that as AF progressed, continued functional and structural changes occurred.⁸ The fibrotic regions contained the largest amount of conduction changes, suggesting that despite global remodeling occurring throughout the atria, the fibrotic regions may have played a more implicit role in sustaining arrhythmias. The fibrotic regions, with slowed conduction and increased unidirectional propagation, may provide the basis of the fibrillatory area for AF to sustain.³¹ However, as remodeling occurs throughout the LA, the true fibrillatory area may be anchored to fibrotic regions but include healthy regions of the atria.

CONDUCTION DYNAMICS.

There are differences in conduction dynamics when stressing the atria at different paced cycle lengths. The alteration in the conduction dynamics as AF progresses illustrates changes to the underlying tissue and may provide an indication of the arrhythmogenicity of the tissue. From the baseline study to the 3-month study, there was a noticeable decrease in conduction speeds at all pacing cycles throughout the atria. The slope of the restitution curve dampened. As AF was sustained for 6 months, the regions of the atria that developed fibrosis experienced further conduction slowing. Consistent with previous reports, there were conduction restitution curve differences between different tissue regions.^16,32

Conduction dynamics may provide some indication of the arrhythmogenicity of the cardiac tissue. Regions with broadened conduction restitution curves may play a role in promoting re-entry and sustaining AF. Previous studies have shown that there are alterations in conduction directions with regions of broad conduction restitution.³³ Other studies have associated rotors with broad restitution curves in the atria and ventricles.^16,34 Our study showed that there were broadened restitution curves over time, and regions that developed fibrosis had smaller slopes and more broad restitution curves compared with healthy regions. The broadening of the restitution curve may indicate progressive structural remodeling.^16,33 Previous studies have observed broad conduction dynamics in patients with persistent AF.³³ Conduction slowing and broadening restitution curves in fibrotic regions may be the result of myocyte-fibroblast coupling, gap junction expression changes, or fibrosis obstructing cell-to-cell conduction.^7,18,35–38

STRUCTURAL REMODELING.

Cardiac LGE MRI and histology samples of the atria provided information on the structural remodeling that occurred as AF sustained for 6 months. Previous works have shown that patients with AF have increased cardiac fibrosis in the atria.^22,23 The cardiac fibrosis development may be due to AF, as AF may increase fibroblast proliferation.^8,37,38 Preclinical models have also shown that animals with fibrosis are more susceptible to sustaining AF.^10,19 Using LGE MRI, our model indicated that fibrosis in the atria increased over time as AF developed and sustained. After a sustained period of AF, there was a notable increase in fibrotic area in the atria. This is consistent with previous works and suggests that as AF persist, the structural remodeling in the atria naturally develops, furthering the disease development.⁸ The fibrosis shown by LGE MRI was confirmed using histologic samples with fibrotic regions, labeled by LGE MRI, having a higher percentage of fibrosis compared with healthy regions.

CLINICAL IMPLICATIONS.

This study furthers our understanding of how structural remodeling affects conduction as AF progresses. Structural remodeling led to sustained conduction slowing and affected conduction propagation directions. The study highlights that increased fibrosis may lead to worsening functional characteristics, and the data suggest that fibrotic regions may provide a structural substrate for a fibrillatory area to form with re-entrant circuits forming from slowed conduction and potential conduction lines of block from unidirectional propagation.

Pulmonary vein isolation has been the standard ablation strategy for treating patients with AF. However, as AF becomes persistent, pulmonary vein isolation has been shown to have less efficacy compared with paroxysmal AF. The mechanism sustaining persistent AF is much more complicated. Our results indicate that functional and structural remodeling occur and continue to remodel further as AF progresses, with the greatest functional changes occurring in the regions that developed fibrosis, but the changes are not focal and are diffuse, distributed throughout the atria. The diffuse changes throughout the atria complicate how we may identify the true fibrillatory area sustaining AF.

Clinical studies such as DECAAF II (Efficacy of Delayed Enhancement MRI-Guided Ablation vs Conventional Catheter Ablation of Atrial Fibrillation) and ALICIA (Isolation of Pulmonary Veins With the Aid of Magnetic Resonance Imaging) have targeted ablating regions of cardiac fibrosis as identified on cardiac MRI.^39,40 However, these trials did not show improved success compared with conventional pulmonary vein isolation ablation. Our data show that though fibrotic regions may have increased arrhythmogenic potential because of their slowed conduction and increased unidirectional propagation, there are diffuse functional changes throughout the LA, so targeting solely the fibrotic area may not improve outcomes as these trials showed. As remodeling occurs throughout the atria, there might not be fixed drivers of AF anchored to specific locations. Moreover, even when existing rotors are successfully ablated, with diffuse and gradual changes throughout the atrium, other areas might provide the necessary conduction slowing and heterogeneity to sustain AF.^11,32 Further research investigating the fibrillatory area may provide refined and improved targets for catheter ablation.^31,41

STUDY LIMITATIONS.

We analyzed and compared the conduction differences in the LA over time. This was done because the animal model involved a pacing lead implanted in the right atrium. The pacing lead led to artifacts in the LGE MRI in the right atrium that prevented structural analysis in the right atrium, limiting the analysis to the LA. We do not consider this to be a major issue, as prior studies have shown that most of the AF is driven by the LA.^42,43

We did not analyze fiber directions.⁴⁴ A comparison between conduction direction and fiber direction could be a focus of a future study. As fiber directions were not obtained, conduction anisotropy measured on true fiber directions was not calculated. Conduction direction heterogeneity was used instead to evaluate propagation directions. Our results of more aligned conduction directions in fibrotic regions supports previous works demonstrating increased anisotropy in AF animals.¹¹

The animals studied were 1 to 2 years of age. This is equivalent to a young adult, which is not the typical age of patients experiencing AF. Despite that limitation, such large animals have long been used to for such studies, as they provide a unique way to study many aspects of the disease as studied here. Also, we had persistent AF in these animals, and it was only the use of such a model that allowed us to pursue this study with serial mapping investigating progression over time and finally comparing it with imaging and histology.

CONCLUSIONS

Functional changes occur over time as AF progresses throughout the atria, with the greatest changes anchored to regions of structural remodeling. Regions that developed fibrosis experienced conduction slowing and increased conduction direction alignment, possibly providing the substrate basis for re-entrant circuits to develop.

PERSPECTIVES.

COMPETENCY IN MEDICAL KNOWLEDGE 1:

Remodeling occurs throughout the atria as AF progresses, and the atria continue further remodeling over time.

COMPETENCY IN MEDICAL KNOWLEDGE 2:

Functional remodeling occurs throughout the LA and is not confined to fibrotic regions. Generally ablating fibrotic regions may not effectively target the fibrillatory substrate.

TRANSLATIONAL OUTLOOK:

Further studies identifying the contiguous fibrillatory area that allows AF to sustain might be necessary as an ablation strategy given the diffuse nature of these structural changes.

ABBREVIATIONS AND ACRONYMS

AF

atrial fibrillation

EGM

electrogram

EP

electrophysiological

ERP

effective refractory period

LA

left atria/atrium

LGE

late gadolinium enhancement

MRI

magnetic resonance imaging