Termination of atrial fibrillation by flecainide: mechanistic insights from spectral analysis

This editorial refers to ‘Increase in organization index predicts atrial fibrillation termination with flecainide post-ablation: spectral analysis of intracardiac electrograms’ by J. Tuan et al ., on page 488. The pathophysiology of atrial fibrillation (AF) is complex and remains incompletely understood. Atrial fibrillation is usually initiated by rapidly discharging foci, often from within the pulmonary veins, which is then maintained by an abnormal left atrial substrate. Experimental evidence suggests that AF is driven by rapidly discharging foci and waves rotating around fixed structures (mother rotors). 1 These drivers give rise to daughter wavelets which propagate around the atria giving the appearance of ‘fibrillatory conduction’. Recurrent paroxysms of AF cause structural and electrophysiological remodelling of the left atrium eventually allowing it to persist. 2 Within hours, there are changes in ion channel expression which shorten action potential duration (APD) and effective refractory period (ERP). 2 Ultrastructural changes beginning within days of onset of AF include left atrial fibrosis, accumulation of extracellular matrix, and reduced expression of connexin 40 reducing conduction velocity. 2 Heterogeneity of refractory periods and zones of slow conduction promote re-entry and support wavelets propagating around the atria. Flecainide exerts its effect through blockade of sodium channels ( INa ) and potassium channels ( IK1 and IKAch ), which reduces conduction velocity and increases APD and ERP, respectively. 3 These changes have opposing effects on wavelength, which is a product of conduction velocity and ERP. An increase in ERP and AF cycle length was demonstrated in response to flecainide in a canine model of AF. 4 A concomitant increase in size and decrease in the number of re-entry wavelets was observed on high-density epicardial mapping. The authors assumed that the prolongation of ERP was the predominant effect, and that re-entry was extinguished by the resultant increase in wavelength and elimination of the excitable gap. However, experimental evidence has not consistently supported this assumption. Direct measurement of wavelength following flecainide administration in a model of AF in goats demonstrated a decrease in wavelength and widening of the excitable gap. 5 The authors proposed that flecainide might slow conduction at ‘pivot points’, where wavefronts make sharp U turns around areas of block, in a similar manner to that observed in a study mapping atrial flutter during administration of moricizine. 6 Reduced ability of wavefronts to follow a course with high curvature may force wavelets to enlarge their circuits (which limits their number), incorporating other areas of functional or anatomical block. The genesis of new propagating wavelets by wave fragmentation near focal drivers may also be reduced. This reduction in the number of re-entry circuits and daughter wavelets would be expected to increase the probability of termination. Computer modelling has been used to examine the effect of sodium channel blockade on AF. 7 This again showed a reduction in conduction velocity and wavelength producing a slowing of re-entry circuits and AF cycle length, and also confirmed an increase in the critical angle of rotation of wavefronts and hence the area needed for them to rotate. This had several key effects, in particular an increase in the core radius of the mother rotors, increasing the probability that they will exhaust the substrate available for depolarization and become extinguished. The tendency for rotors to meander was also increased, which increases the risk of collision with electrically inert structures such as the mitral annulus or pulmonary veins. Wavefronts also had a decreased propensity to fractionate, producing fewer daughter wavelets. These insights into the mechanisms of AF come mostly from animal models in dogs and goats, and subtle inter-species variation in the cellular effects of flecainide may change the mechanism of action. Although studies on humans confirm the reduction in conduction velocity and increase in APD and ERP with flecainide, 3 , 8 confirmation of the proposed mechanism of action on a whole organ level is awaited. Studying mechanisms of AF in vivo and the response to antiarrhythmic drugs and ablation are limited in part by the resolution of conventional mapping techniques. The rapidly discharging foci and rotors described experimentally are thought to be only 2–3 mm across. 1 The resolution of conventional in vivo mapping techniques is realistically limited to 5–10 mm. It has been proposed that fractionated signals (i.e. continuous deflections in a prolonged activation sequence) on bipolar electrograms represent areas bordering these foci where conduction fails to propagate in a 1:1 fashion. 1 Although electrograms are conventionally examined in the ‘time domain’, with time on the x -axis and signal amplitude on the y -axis, chaotic signals and electrogram fractionation make mapping and establishing mechanisms difficult. This has prompted signal analysis in the frequency domain, where the electrogram is separated into its constituent signals of different cycle lengths and displayed with frequency on the x -axis and amplitude on the y -axis. The signal with the highest amplitude is said to be the dominant frequency (DF), and it is thought that in a chaotic signal, this may represent the frequency of depolarization of the tissue in closest proximity to the catheter electrodes. The organization index (OI) is defined as the area under the DF peak and its harmonics divided by the total area under the power spectrum curve, and reflects the proportion of tissue depolarizing at that frequency. The OI may also indicate the number of different drivers since each has the potential to cause a separate peak in the frequency spectrum, hence reducing the OI. In the study by Tuan et al ., 9 flecainide was administered to patients following catheter ablation of AF and intracardiac electrograms analysed in the time and frequency domain for insights into the mechanism of action. The authors demonstrated an increase in AF cycle length and a corresponding reduction in DF after administration of flecainide. There was an increase in OI in patients who reverted to sinus rhythm (measured immediately prior to restoration of sinus rhythm), although there was no change in patients who remained in AF. Flecainide restored sinus rhythm more often in those with paroxysmal AF (6 of 16) than those with persistent AF (1 of 10). The increase in AF cycle length and decrease in DF observed following flecainide administration is compatible with the slowing of conduction velocity seen with sodium channel blockade, and may reflect slowing of the driving rotors or re-entry circuits. Although this may contribute towards the restoration of sinus rhythm, there was no difference in response between those who cardioverted and those who did not, suggesting either that there were mechanisms other than slowing of drivers causing termination of AF or simply that this was sufficient to terminate AF in some patients. Increased OI was observed only in those in whom sinus rhythm was restored, suggesting a decrease in the number of drivers and other active re-entry circuits such as wandering wavelets in these patients. However, the difficulty in confirming changes in wavelength or excitable gap means there remains uncertainty as to the exact mechanism of action. Increased APD and ERP causing an increase in wavelength and reduction in the excitable gap or slowing of conduction velocity at pivot points would both explain these findings. The weight of data from animal models supports the later mechanism. 5 , 10 Studies in humans also suggest that flecainide has only a minor effect on APD and ERP. 3 , 8 The slowing of cycle length and reduced DF demonstrated strongly suggest slowing of conduction velocity, making this the more likely mechanism of termination of AF by flecainide in these patients. The low proportion of patients with persistent AF who cardioverted with flecainide suggest that the remodelling process either diminishes the electrophysiological effect of flecainide or increases the stability of re-entry circuits rendering them resistant to the effects of flecainide. Evaluation of the electrophysiological effects of flecainide before and 48 h after induction of AF in goats showed similar results, suggesting that it is the increasing stability of re-entry circuits with remodelling of the atria that causes resistance to flecainide. 10 The failure of OI to increase in those who did not cardiovert may also point to advanced atrial remodelling in these patients, with greater stability of re-entry circuits explaining the resistance to flecainide. Although persistent AF was identified as an indicator of failure to cardiovert following flecainide administration, no electrophysiological parameter recorded prior to administration predicted response. Hence, although this study adds to our understanding of the mechanism of AF termination by flecainide, there remains a missing element to our understanding of how atrial remodelling renders the multiple re-entry circuits resistant to the action of flecainide. Conflict of interest: none declared.