Atrial inputs as determinants of atrioventricular nodal conduction: re-evaluation and new concepts

Radiofrequency ablation was combined with standard recording and pacing procedures to determine the role of the perinodal fast pathway (FP) and slow pathway (SP) in atrioventricular (A-V) conduction in normal dog hearts. In 16 anesthetized, open-chest dogs, we recorded atrial (A) electrograms from the high right atrium (HRA), His bundle (Hb) region and coronary sinus (CS) ostium. In 8 dogs, during HRA pacing at a cycle length of 400 ms, FP ablation induced a significant change in A-H interval (61 +/- 12 to 107 +/- 9 ms, p < 0.001) but no change in Wenckebach cycle length (185 +/- 14 to 190 +/- 12 ms, NS). During ventricular pacing (400 ms cycle length) there was a shift in retrograde atrial activation so that Hb (A) delayed more than CS (A). Subsequent SP ablation in this same group of 8 dogs, produced a significant prolongation of antegrade Wenckebach cycle length (190 +/- 10 to 277 +/- 36 ms, p < 0.002) but only one instance of complete A-V block even though both FP and SP were ablated in accordance with clinical criteria. In another group of 8 dogs, a specially designed ablation electrode was inserted into the right atrium and positioned at the level of the fossa ovalis, 10-12 mm superior to the perinodal area. A radiofrequency linear lesion across the interatrial system induced a marked PR interval prolongation due to an intra-atrial rather than A-V nodal delay. The A-H interval was unchanged, and there was no change in antegrade or retrograde Wenckebach cycle length. In contrast, to the findings in both these groups direct damage to the A-V node showed immediate change in the A-H interval and at the same time progression to second degree and complete heart block. Taken together these results suggest that the perinodal FP and SP are composed of transitional cells which possess distinct electrophysiological properties not shared by adjacent atrial tissues or the compact A-V node itself. We conclude that the persistence of A-V conduction, albeit modified, after FP and SP ablation, suggests the existence of multiple atrio A-V nodal inputs whereas retrograde conduction relies mainly on dual exits from the A-V node to the atria.