Structural Remodeling in Atrial Fibrillation as Arrhythmogenic Substrates: Tachyarrhythmia and Gap Junction Remodeling

Tomoko Ohkusa
Yamaguchi University School of Medicine, Ube-Yamaguchi, Japan

Fibrosis, extracellular matrix, fiber orientation, autonomics, Gap junction, ion channels, and calcium homeostasis are known to be arrythmogenic substrates. Remodeling is thought to trigger arrhythmias and facilitate their occurrence, which can induce further remodeling. The downregulation of the ryanodine receptor or Ca²⁺ ATPase and the upregulation of IP3 receptors might cause an abnormal intracellular calcium homeostasis, resulting in the initiation and perpetuation of atrial fibrillation (AF).

Gap junction remodeling

Electrical coupling in the heart is mediated by Gap junctions, specialized membrane lesions comprising groups of channels directly connecting the cytoplasmic compartments of two adjacent cells. Gap junction facilitates the intercellular exchange of ions, signaling molecules, and other molecular information, and affects impulse propagation and/or electrical coupling (Figure 1). In the heart, at least 4 different connexin (Cx) proteins are expressed, with short half-lives of 1-2 hours. Phosphorylation regulates the stability of the connection.

Chronic fibrillating human atrial myocardium

This group analyzed the expression of Cx40 and Cx43 in the right atrium of patients with mitral valvular disease (MVD) with AF or in normal sinus rhythm (NSR) and in controls. Cx40 mRNA expression was significantly lower in MVD/AF patients, while no significant difference was seen in for Cx43 between the 3 groups. Expression of Cx40 protein was significantly lower in AF patients, while for Cx43 protein there was no significant change. Serine-phosphorylation, but not tyrosine phosphorylation, was found for both Cx40 and Cx43 on tissue immunoprecipitation. Notably, serine-phosphorylated Cx43 was about 50% greater in patients with AF.

Cardiomyocytes exposed to rapid electrical stimulation (RES) of contraction

Cx43 mRNA was increased in rat cultured ventricular myocytes, peaking at 90 minutes. Cx43 immunoreactive signals were appreciably increased with no affect on the distribution pattern (Figure 2). Quantitative confocal analysis revealed a significant increase in the Cx43 area.

RES was associated with a marked increase in angiotensin II content in the cultured cardiomyocytes (Figure 3). Angiotensin II is known to increase Cx43 expression. However, in presence of losartan, RES did not significantly increase Cx43 protein or Cx43 mRNA.

Nearly uniform propagation of excitation from left to right was seen on multielectrode extracellular potential mapping. In the cultured cardiomyocytes, compared to controls, the activation times at the recording sites after RES were much shorter—indicating a RES-induced increase of propagation, which did not occur in cardiomyocytes treated with losartan. In the absence of losartan, RES increased the conduction velocity after 30 minutes, while no significant changes in conduction velocity by RES were seen in the presence of losartan.

Cardiomyopathic hearts

In hamster cardiomyopathic hearts, cell width and percentage of fibrosis gradually increased after 10 weeks. After 20 weeks, left ventricular end diastolic diameter (LVEDD) significantly increased, and the ejection fraction (EF) and fractional shortening decreased in the cardiomyopathic hamsters compared to age-matched control hamsters. QRS width was significantly prolonged at 20 weeks (Figure 4). Quantitative analysis of Cx43 revealed that the amount of Cx43 mRNA and Cx43 protein were decreased at 20 weeks. In the cardiomyopathic hamsters, Cx43 expression on immunohistochemistry was gradually decreased compared to golden hamsters (Figure 5). Quantitative analysis of serine 255 phosphorylation of Cx43 showed its expression was gradually decreased with age in cardiomyopathic and control hamsters. Interestingly, in the cardiomyopathic hamster, the expression of serine phosphorylated Cx43 was significantly greater than that of golden hamsters at 6 weeks and 20 weeks.

Summary

In chronically fibrillating human atrial myocardium, Cx40 was downregulated and abnormally phosphorylated, but no changes were seen for Cx43.
In ventricular cardiomyocytes exposed to rapid electrical stimulation of contraction, Cx43 was upregulated through an autocrine action of angiotensin II to active MAP kinases, resulting in potentially arrhythmogenic alteration of conduction properties.
In cardiomyopathic hamster hearts, QRS interval was prolonged with an increase of interstitial fibrosis and downregulation and abnormal phosphorylation of Cx43. 

Conclusion

In addition to functional remodeling of cardiomyocytes, gap junction remodeling might also cause abnormal impulse propagation and electrical coupling, resulting in the initiation and/or perpetuation of arrhythmias. For new therapeutic target, the upstream approach to regulating substrates involved in remodeling at the molecular, cellular and/or organ level may be a promising new approach in the treatment of tachyarrhythmias, including AF.

Effects of Antiarrhythmic Drugs on Electrophysiological Action and Wavefront Dynamics During Atrial Fibrillation

Takanori Ikeda
Kyorin University School of Medicine, Mitaka, Japan

The wavelength theory, which states that the prolongation of the wavelength (WL), either of the refractory period (RP) or increment of conduction velocity (CV), has been used to explain the efficacy of antiarrhythmic drugs to terminate AF. However, the WL theory does not fully explain the mechanism of action whereby all antiarrhythmic drugs terminate AF.

Recently, the importance of the “temporal” and “spatial” excitable gap (EG) for cardioversion of AF has been proposed by Wijffels and colleagues. The temporal EG, which is thought to be more important, is calculated as the mean AF cycle length minus the refractory period. The spatial EG is calculated as the pathlength minus the wavelength. They state that the temporal and spatial EG are widened, but the WL not prolonged, just before cardioversion of AF with multi-channel blockers, such as cibenzoline and flecainide.

The single meandering hypothesis and the mother rotor hypothesis have shown that wavefront propagation is complex, but an essential reentry exists during AF. It has also been shown that the property of the core is important for the maintenance of functional reentry in the atrium, and the reentry terminates when the core is excited by an outside wavefront.

Study design

To determine alternative mechanisms by which antiarrhythmic drugs terminate AF, they hypothesized that a widening of the EG, the enlargement of the core, and the presence of outside wavefronts participate in cardioversion of AF by antiarrhythmic drugs.

To test their hypothesis they studied pilsicainide (PIL), a pure sodium channel blocker, whose high efficacy cannot be understood because its main action is the decrement of conduction velocity, and nifekalant (NIF), a pure Ikr channel blocker, whose lack of efficacy in AF is not understood, despite its ability to prolong the WL and RP and its efficacy in ventricular fibrillation.

Mapping techniques were used to assess the effects of PIL and NIF on electrophysiologic actions, including the EG, and wavefront dynamics during AF. In a newly developed model of isolated, perfused and superfused canine atria, AF was induced with 1-5 mM acetylcholine (Figure 1).
The right and left endocardia were simultaneously mapped using a computerized mapping system.
 
 PIL (2.5 mg/mL) or NIF (5 mg/mL) was perfused after stabilization of induced AF.
Electrophysiological parameters and the core of mother reentry (an essential reentry) were measured during AF. An electrode array with 224 bipolar electrodes was used. The temporal EG was calculated as the mean AF cycle length (ms) minus the RP (ms).

Study results

After perfusion of PIL, the AF cycle was increased 63%, the RP increased 41%, the CV velocity decreased 40%, and the WL decreased 15% (Figure 2). A significant increase in the EG and core area were found (147% and 100%, respectively). In 8 of 12 preparations, AF was eventually terminated with a cumulative PIL dose of 7.5 É g/mL. In these 8 preparations, the activation pattern was unorganized, despite the decrease in the number of WL (Figure 3). AF was converted to an organized pattern in the remaining 4 preparations.

With NIF, the activity was regular and uniform on the bipolar electrograms. The reentrant wavefront rotated in a counter clockwise direction. After NIF perfusion, only a single reentrant wavefront can be seen. The mean cycle length was increased 67%, the RP increased 15%, the CV 21%, and the WL 21% (Figure 4).  A marked widening of the EG and enlargement of the core perimeter were found (289% and 155%, respectively).

Although NIF widened the EG and the core perimeter, no outside wavefronts were observed. Further experiments showed that with NIF organized AF was easily terminated by a single pacing from the EG area. Thus, an outside wavefront appears necessary to terminate AF.

Conclusions

Outside wavefronts, widening of the EG, and enlargement of the core of the mother reentry are important for cardioversion of AF by pure Na channel and Ikr channel blockers, according to the data from the present study.

Long-Term Prognosis of the Patients with Paroxysmal Atrial Fibrillation: Differential Responses to Antiarrhythmic Therapy

Takashi Komatsu
Iwai Hospital, Ichinoseki, Japan

A retrospective analysis of patients with paroxysmal atrial fibrillation (AF) by Komatsu and colleagues showed a variable long-term prognosis related to the response to the antiarrhythmic therapy. Maintenance of sinus rhythm (SR) was associated with a good prognosis, even in the absence of anticoagulation therapy. AF recurrence or development of permanent AF was associated with a poor prognosis, with the possibility of stroke in the absence of anticoagulation therapy. Personalized selection of antiarrhythmic drugs is needed to improve its efficacy. The development of a more effective, more atrium-specific and safer drug is anticipated.  

Study design

To answer questions remaining after the AFFIRM and RACE trials, these investigators sought to clarify the role of antiarrhythmic drug therapy in the management of patients with symptomatic paroxysmal AF and with normal or minimally impaired left ventricular (LV) function. They examined the 1) long-term efficacy of antiarrhythmic drug therapy with class I drugs and amiodarone, and 2) relationship between the efficacy of antiarrhythmic drug therapy and the long-term prognosis of the patients.

A total of 290 patients with symptomatic paroxysmal AF (191 men, mean age 69 years) and an LV ejection fraction (EF) > 40%, who were treated þ 1 a month at the outpatient clinic for a minimum of 12 months (mean follow-up 51 months) was included in the analysis. According to the protocol for antiarrhythmic therapy, after restoration of sinus rhythm (SR) a class 1 drug was given by a sealed envelop method. For AF recurrence, either flecainide, pilsicainide, or bepridil was given by a sealed envelope method. If AF persisted, the physician could freely select amiodarone or a different class I drug. Figure 1 illustrates the protocol for drug therapy in this study.

Study results

The efficacy of the antiarrhythmic drug in preventing the recurrence of paroxysmal AF is limited. At 20 months after the first antiarrhythmic drug was given, only 51% of the patients on disopyramide, 47% oncibenzoline, and 35% on aprindine were free of AF recurrence. At 20 months after the second antiarrhythmic drug, only 33% of patients treated with flecainide and with pilsicainide and 21% with bepridil remained free of AF recurrence. After the addition of the third antiarrhythmic drug, at 20 months, 44% of patients treated with amiodarone and 60% with a class I antiarrhythmic drug remained free of AF recurrence.

Selection of the antiarrhythmic drug based on the clinical profile of the patient may improve efficacy. A circadian variation has been shown in the paroxysmal AF in many patients. The investigators divided the patients taking disopyramide into 3 groups based on the time of AF recurrence: diurnal (n=14), nocturnal (n=34), and mixed (n=37). At 24 months after the initiation of disopyramide, 71% of the nocturnal group was recurrence-free, compared to only 43% of the diurnal group and 27% of the mixed group (Figure 2). Therefore, disopyramide can be first-line treatment for patients with nocturnal type of paroxysmal AF without contraindications.

Based on the results of the antiarrhythmic therapy, the patients were divided into 3 groups: SR without recurrence (Group 1, n=114), AF recurrence during therapy (Group 2, n=113), and conversion to permanent AF (Group 3, n=63). The efficacy rates of 39%, 39%, and 22%, respectively, are consistent with other reports, showing the limited efficacy of antiarrhythmic drugs to maintain SR.

At 16 months, no significant difference in survival free of cardiovascular death was found between the 3 groups (99%, 95%, and 94%, respectively). Survival free of fatal or nonfatal symptomatic ischemic stroke was significantly higher in Group 1 compared to Groups 2 and 3 (99%, 88%, and 76%; p<0.05; Figure 3). Ischemic stroke occurred in only 1 patient in Group 1. The incidence of ischemic stroke in patients with AF clearly depends on the antithrombotic therapy (Figure 4). At baseline, no significant difference was noted in the prevalence of the risk factors for ischemic stroke among the 3 groups. The total annual stroke rate was lower in Group 1 than in Groups 2 and 3. In patients in whom any antithrombotic therapy was not given during follow-up, the annual stroke rate was lower in Group 1 than in Groups 2 and 3. In patients in whom antithrombotic therapy was given the annual stroke rate was higher in Group 1 than in Groups 2 and 3. In patients taking aspirin, the annual stroke rate was higher in Group 1 than in Groups 2 and 3, while in the patients taking warfarin the annual stroke rate was similarly low in each group.

Map-Guided Surgery for Atrial Fibrillation

Takashi Nitta
Nippon Graduate School of Medicine, Tokyo, Japan

The Maze procedure cures atrial fibrillation (AF) by isolating all four pulmonary veins (PVs) and by the interrupting all possible reentrant circuits. Although the success rate of the Maze procedure is 90%, it is a technically demanding procedure, involving complex and extensive atrial incisions, bleeding risk, 40 to 50 minutes of cardiac arrest time often required. Recently, isolation of the PV alone has successfully cured AF in selected patients. The PV isolation procedure is simple, safe, and can be completed within 10 to 15 minutes. However, the success rate of 50-70% for permanent AF is significantly lower than with the Maze procedure. Electrophysiologic-based surgery that is less invasive needs to be established to address some of the problems remaining with the current procedures.

Study design

To address this need, this group performed intra-operative mapping in 45 patients (25 male). The average age was 62 years, 82% had valvular heart disease, 4 patients had congenital heart disease, and 4 patients no structural heart disease. Paroxysmal AF was found in 9 patients and 36 had permanent AF.

Custom-made electrodes were used for the intraoperative mapping. An atrial mold from a cadaver heart was made, and then the curvature of the atrial epicardium was copied to 3 silicon patches, one each for the lateral right atrium, the Bachmann’s bundle, and the lateral left atrium. A total of 253 bipolar electrodes were distributed over the 3 silicon patches. Nearly the entire epicardium of the atria, including the posterior left atrium between the pulmonary veins, was mapped with the electrodes.

AF of 500-1000 ms duration was analyzed in each patient. Activation sequence maps were displayed on a 3-D atrial model as a movie. At least 10 minutes were required for the analysis and more time was required in complex activations.

Findings from intra-operative mapping

A representative map from a 72-year-old male patient with mitral valve regurgitation, coronary artery disease (CAD), and paroxysmal AF revealed at least 2 foci of repetitive activation—one in the left superior PV (LSPV) and the other is in the right superior PV (RSPV). Propagation was toward the right atrium and over the Bachmann’s bundle, because of the faster activation from the LSPV than from the RSPV. The right atrium activation was passive in this patient, and he was cured of AF by PV isolation alone, without incisions on the right atrium, because there was no reentrant activation in the RA.

In a patient with permanent AF and mitral valve disease, focal activation originated from the LSPV and propagated toward the right atrium and over the Bachmann’s bundle. Another focal activation originated from the RSPV simultaneously. Left to right inter-atrial conduction via Bachmann’s bundle demonstrated so-called fibrillatory conduction. The electrograms revealed a progressive conduction delay in the pathway from the focus at the LSPV to the right atrium, resulting in a variable conduction ratio and non-activated regions.

All patients had 2 to 4 foci in the left atrium, mostly arising from the posterior left atrium adjacent to the PVs, particularly the LSPV and RSPV. Interestingly, this distribution is similar to that in paroxysmal AF shown by Haissguerre and colleagues.

In the passive activation on the lateral right atrium during AF, the focal activation arose from the PV, conducted over the Bachmann’s bundle, and appeared at the top of the right atrium (Figure 1). The activation in the right atrium was passive, with progressive conduction delay in the lateral right atrium. The delayed conduction resulted in residual activation from the preceding cycle and coexisting multiple wavelets in the lateral right atrium. As a result, the activation in the right atrium desynchronized with the left atrial activation, and became irregular and complex.

Concurrent multiple PV focal activations combined with fibrillatory conduction is the dominant mechanism in the basic pattern of atrial activation during AF (Figure 2). Extremely rapid activations in the PV and a progressive conduction delay or block in the pathway from the focus to the right atrium and in the lateral right atrium cause an irregular and complex right atrium activation, resulted in coexisting multiple wavelets.

The map from a 68-year-old male patient with mitral valve stenosis, tricuspid valve regurgitation, and chronic AF revealed focal activation arising from the LSPV, and a reentrant activation around the right atrial appendage (Figure 3). The patient required the radial procedure comprising biatrial incisions in addition to PV isolation.

More than 50% of the study patients with permanent AF had right atrium activation, but none of the paroxysmal AF patients.

In 3 patients, the electrograms were low voltage over the broad area in the left atrium. There was a focal activation arising from the LSPV, and the right atrium was activated by the activation conducted from the left atrium (Figure 4). Surgery for AF was not indicated, because of the low chance of converting AF to sinus rhythm, and the poor atrial contraction even if sinus rhythm resumed.

Strategy for intra-operative mapping guided surgery

Treatment approaches based on the findings from intraoperative mapping are as follows. For PV focal activations and passive conduction in the right atrium, cure may be achieved by PV isolation alone. For right atrial reentrant activations, the Maze or radial procedure should be used to interrupt the reentrant circuits. For low voltage electrograms in a broad area, particularly in the left atrium, no surgery is performed, because of the low rate of converting AF.   

Study results

Of 45 patients, 3 had low voltage electrograms in the broad left atrium area, and AF surgery was not indicated (Figure 5). In the remaining 42 patients, 8 patients clearly demonstrated passive activation in the right atrium combined with focal activations in the left atrium. The PV isolation was performed in these patients, and sinus rhythm resumed in all. In the remaining 34 patients, reentrant activation was observed in the right atrium and focal activation in the left atrium. The radial procedure was performed, with a 91% success rate. 

Conclusions

Intra-operative mapping is useful to determine the optimal procedure in each patient.
In patients with passive activation in the right atrium without reentrant or focal activations, simplified procedures confined to the left atrium or the PVs may be indicated. However, ablation of a single focus may fail to cure AF because there can be more than one source of focal activation and the second focus would assume dominance after the ablation of the primary focus.

The limitations for the present system for intra-operative mapping include the lack of data from the septum, the effect of anesthesia, the possibility that the activation pattern may vary, and the considerable time to analyze the data and determine the activation pattern. A solution may be pre-operative catheter mapping. Right atrial endocardial mapping using a basket catheter or non-contact balloon catheter electrode may determine whether the right atrial activation pattern is passive or reentrant. Within five years, it is likely that mapping or a electrophysiologic-guided procedure will be established. The procedure will be minimally invasive, probably using a robot without the use of cardiopulmonary bypass.