Background: Catheter based "maze" procedures for atrial fibrillation have been hampered by difficult creation and evaluation of continuous and transmural linear lesions. Our aim was to develop an online evaluation method for effective lesion creation based on conventional techniques and using the multipolar ablation catheter, already in place.
Methods and results: We created 15 linear lines in right atria of 13 anesthetized sheep using three multipolar catheter designs (8 x 4 mm 7 Fr, 4 x 6 mm 7 Fr, 8 x 4 mm 3.7 Fr). The lesions were placed on the right posterolateral wall between the orifices of the superior and inferior vena cava. Radiofrequency energy was applied in the temperature-controlled mode to prespecified endpoints (electrogram amplitude decrease to < or = 50%; pacing threshold increase by > or = 100%; split potentials indicating conduction block). Macroscopically transmural and continuous lesions were achieved in only 3 experiments (29 +/- 12 mm x 5 +/- 1 mm), all created by 3.7 Fr octapolar catheters inserted through long sheaths. Preset temperature was reached in 96% of the electrodes (vs. 64% in the non-effective experiments; p < 0.01). Electrogram amplitude decrease (to < or = 50%) and pacing threshold increase (by > or = 100%) did not predict effectiveness. The only criterion that could reliably predict transmural continuous necrosis at histology was the development of split potentials (p < or = 0.05).
Conclusions: Effective creation of linear lesions is difficult. Pliable catheters that conform to the endocardial contour give the best results. The only endpoint that reliably predicted histological transmural continuous necrosis was development of split potentials indicating conduction block.