Surgical Technique
Re-sternotomy was performed in all cases. The femoral artery and vein were prepared in cases where the mediastinal structures were adhered to the sternum (determined with lateral chest x-ray or computed tomography in the preoperative evaluation), whereas femoral cannulation and initiation of cardiopulmonary bypass before sternal entrance was not necessary in any case. Following aortic and bicaval (or tricaval in cases with a persistent superior venae cavae), aorta was cross clamped and antegrade blood cardioplegia was administered where intermittent doses were repeated at every 15 minutes. Oblique right atriotomy was performed at the lower half of the atrial wall. Pulmonary arteries were divided above the pulmonary valve except for the cases with pulmonary atresia. Pulmonary valves were either resected or squeezed within the polypropylene double row suture line that will close the pulmonary outflow of the ventricle. If necessary, AV valves were repaired at this stage of the operation. The PTFE conduit with proper size was anastomosed to the orifice of IVC in a beveled end-to-end fashion, carefully observing any accessory hepatic venous drainage to the right atrium. In patients no. 7 and 17, at least one hepatic vein was separately draining into right atrium, which were included into Fontan circulation with the conduit involving their ostium. Afterwards, an aortic punch was used to fenestrate the conduit at the posterior side, just above the IVC anastomosis. Then the graft was gently pulled outside the atrium in order to provide a straight and unkinked part to be secured inside the atrial cavity. The right atrial incision was closed with continous polypropylene suture that will enwrap the conduit at the point where it will be directed out to the pericardial cavity. The pulmonary arterial anastomosis of the graft was performed with polypropylene or PTFE sutures. In cases where a pulmonary arterial augmentation was deemed necessary, a bovine pericardial patch (pt. no 12) or a PTFE patch (pt. no 15) was used. There was no chance to prepare autologous pericardium for this purpose due to mediastinal adhesions in these redo cases. Total circulatory arrest was not used in any case. In patient 12, a MAPCA was occluded with percutaneous intervention one day before the operation. This patient was the only case where a cavopulmonary anastomosis and IECF were performed at the same stage. Temporary inotropic agents were infused in pt. no 12, 13 and 17. Permanent pace maker implantation was not deemed mandatory in any case.
The mean cardiopulmonary bypass and aortic cross clamp times were 93.1 ± 34.2 and 70.2 ± 23.9 minutes, respectively. All of the patients were extubated after 7.2 ± 2.6 hours at the intensive care unit (ICU). The mean ICU stay was 1.7 ± 0.9 days. The chest tubes were kept in place until the daily drainage was less than 2cc/kg. Mean duration of drainage was 5.4 ± 2.3 days. Oral warfarin and aspirin were routinely administered in all of the patients. Steroids were administered when the daily drainage exceeded 5cc/kg after 5 days of follow up. The target prothrombin time (international normalized ratio, INR) was 2 to 2.5. We did not make any second intervention for accumulation of pleural fluid after the tubes were removed, except for patient no.15, whom we had to insert pericardial and pleural drainage tubes via subxiphoid approach 13 days after IECF. We did not encounter any phrenic nerve paralysis in this patient population.
Mean follow up period was 3.6 ± 1.9 years (range: 1 to 7 years). The mean systemic arterial saturation level at room air was 91.3 ± 2.7 % (range: 89% to 96%) at the follow up. In pt. no 11, fenestration was closed via femoral access one year after the IECF. Except this case, all of the fenestrations are functioning during the follow-up period, providing a right to left shunt with 5.2 ± 1.3 mmHg shunt gradient calculated at transthoracic echocardiography. The patients were discharged with diuretics and angiotensin converting enzyme inhibitors (when necessary) in order to achieve systemic decongestion and normal blood pressure.
Fontan failure, associated clinical scenarios (plastic bronchitis or protein losing enteropathy) and mortality was not encountered in our patient population.