br Panels A to D show intracardiac recordings of
Panels A to D show intracardiac recordings of the HRA extrastimuli with A1A2 intervals of 320−290ms at a basic CL of 600ms. The AV conduction intervals were increased gradually rather than suddenly, so that an Ae with an A2–V2–Ae sequence could be provoked only at 310ms and 300ms, followed by no AVNRT initiation. The outcomes of extrastimuli, including other A1A2 intervals, are illustrated as a post-RFA AV conduction curve in Fig. 4A. Panel E shows that the elimination of the retrograde VA conduction by RFA was confirmed during incremental pacing at a CL of 600ms from the RV. However, panel F shows that the VA conduction reappeared with a long His-atrial time (220ms, not shown in the figure) during the AJR provoked by the administration of isoproterenol (Isp). From the unified standpoint, the mechanism of the Ae with the A2–V2–Ae sequence induced after RFA should be ascribed to an atypical fast–slow AVN reentry. Therefore, the retrograde slow AVNP was considered as the native pathway rather than the fast pathway modified by RFA. It was suggestive of the presence of the latent retrograde dual AVNP physiology. Fig. 4A shows antegrade AV conduction curves before (solid line) and after (broken line) RFA without Isp infusion, where the ordinate and abscissa indicate an AV interval (ms) and an A1A2 interval (ms), respectively, the former showing the presence of antegrade triple AVNPs, the latter showing shortness of the effective refractory VE-822 and facilitated conduction of the fast pathway, and elimination of the slow and very slow AVNPs. Judging by these findings, sites impaired by ablations could be determined to include the antegrade slow and very slow AVNPs, and the retrograde fast AVNP but not the retrograde slow AVNP (Fig. 4B) . The programmed stimulation from the HRA performed with Isp infusion after RFA delineated a smooth AV conduction curve and did not produce Ae because of further facilitation of conduction over the antegrade fast pathway (not shown in the figure).
Discussion The study has several limitations. First, since the CL of a clinical PSVT differs from that of an induced AVNRT (400 vs. 510ms, respectively), their mechanisms may differ as well. However, the difference can be explained by different autonomical balances in the clinical state and in our cardiac laboratory. Second, EPS was not performed under Isp administration before the RFA. If the procedure had been done, improvements in retrograde fast and antegrade slow AVNP conduction might have veiled some of the above-mentioned rare electrophysiological phenomena . Third, the hypothesis that the AV conduction curve at baseline implied the antegrade dual AVNP physiology with a longitudinal dissociation of the lower final common pathway instead of the triple AVNP may be accepted. However, since the real potential of the AV node cannot be recorded directly, it is impossible to determine whether the hypothesis is correct. Some of the above-mentioned unique phenomena, including the DVR, would be more advantageous to the triple than to the dual pathway physiology. Finally, although effective pacing methods to evaluate the presence of an upper or lower final common pathway have been proposed [1,4,12], accuracy can be compromised in cases such as ours, which involve the upper and lower final common pathways.
Introduction Atrial fibrillation (AF) is a common cardiac arrhythmia and may be caused by obstructive sleep apnea (OSA). A previous report showed that AF occurred in 4.8% of the sleep-disordered breathing (SDB) group (N=228), whereas only 0.9% of those without SDB presented with AF (p=0.003) . Other common causes of AF include hyperthyroidism, hypertension, and valvular heart disease, particularly mitral stenosis . AF can be a contributing factor for various diseases such as cerebral or systemic embolism . Conversion of AF is crucial to reduce cardiovascular morbidity and mortality [4,5].