A canine model of atrial flutter following the intra-atrial lateral tunnel Fontan operation

Atrial flutter (AFL) is a common problem in children who have undergone a Fontan operation for single ventricle physiology. Although this has been attributed to the atrial stretch inherent in the earlier forms of this operation, AFL has persisted in spite of a modification that minimizes atrial distension. Therefore, it was hypothesized that AFL following the modified Fontan procedure may result from anatomic barriers related to suture lines rather than from atrial stretch or hypertension. In a series of experiments performed in dogs under general anesthesia, the modified Fontan repair was simulated by placing only the suture line of the intra-atrial repair. No baffle was placed, thus avoiding any hemodynamic alterations. After closure of the atriotomy, 253 point unipolar atrial endocardial form-fitting electrodes were inserted through the mitral and tricuspid valves via bilateral ventriculotomies. Induction of AFL was attempted with atrial burst pacing and programmed extrastimulation, and activation sequence maps of subsequent reentry were generated from the endocardial electrodes. Atrial flutter was induced in all of 17 dogs, with a median cycle length of 177 +/- 31 ms. Activation sequence maps demonstrated conduction block along the crista terminalis corresponding to the free wall portion of the suture line. This created an isthmus between the suture line and tricuspid annulus, which appeared critical for sustaining AFL, although the circuit used both the septal and free wall surfaces of the right atrium. In seven dogs, a cryolesion was placed from the tricuspid annulus to the free wall segment of the suture line, terminating the AFL, in all seven. When the free wall segment of the suture line was moved 5 mm medial to the crista terminalis, AFL was induced in four of five dogs, but only in the presence of isoproterenol and at a shorter cycle length (136 +/- 8 ms, P < .001). Atrial flutter was not inducible, even with the addition of isoproterenol, in any of five dogs in which the suture line was placed 10 mm anterior to the crista terminalis and incorporated into closure of the atriotomy. This acute canine model of the modified Fontan operation demonstrates that conduction block from the free wall portion of the suture line creates an isthmus of tissue between the suture line and the tricuspid annulus. This is a sufficient substrate to produce AFL; no hemodynamic alteration is required. Injury to the crista terminalis is a significant risk factor in this model, which suggests that a modification of the suture line might reduce the incidence of AFL in patients following this operation.