Objective: Neonates with profound heart failure resulting from Ebstein anomaly have historically had poor outcomes. We report our institutional experience with the surgical management of Ebstein anomaly in severely symptomatic neonates.
Methods: A retrospective review of all patients (n = 16) undergoing neonatal intervention for Ebstein anomaly between 1992 and 2005 has been carried out. The indications for operation were overt heart failure, cyanosis, and acidosis associated with tricuspid regurgitation, depressed right ventricular function, and severe cardiomegaly. The magnitude of cardiac enlargement was assessed by cardiothoracic ratio and Great Ormond Street ratio (area of right atrium + atrialized right ventricle/area of functional left atrium + left ventricle). The operative strategy was first to assess for the possibility of tricuspid valve repair with or without right ventricular outflow tract reconstruction. If this was not feasible, then right ventricular exclusion was performed by oversewing the tricuspid valve with a pericardial patch. A reduction atrioplasty was done and, depending on the extent of the atrialized portion of the right ventricle, plication was performed. A modified Blalock-Taussig shunt provided pulmonary blood flow. This univentricular approach (Starnes procedure) evolved to include a fenestration in the tricuspid valve patch to allow for right ventricular decompression. Analysis included overall and group-specific survival as well as the testing of perioperative clinical, morphologic, and surgical variables for correlation with mortality and morbidity.
Results: Mean age and weight at operation were 8 +/-10 days and 3.1 +/- 0.4 kg. Tricuspid valve repair was undertaken in 3 patients with 1 requiring conversion to right ventricular exclusion 3 months after the initial operation. In those with right ventricular exclusion, the tricuspid valve patch was fenestrated in 10 and nonfenestrated in 3. One patient had heart transplant as the initial procedure. There were 5 hospital deaths (31%) and no late deaths among the survivors. Survival in the cohort with a fenestrated tricuspid valve patch was 80% (8/10) versus 33% (1/3) for the nonfenestrated group. This difference did not reach statistical significance, although the trend seems clinically important. There was no difference in the cardiothoracic ratio (0.82 fenestrated vs 0.84 nonfenestrated: P = .802) or the Great Ormond Street ratio (1.2 fenestrated vs 1.02 nonfenestrated: P = .477) between the two groups. Among the 9 survivors of right ventricular exclusion, 3 have had completion of their Fontan, and all 9 have undergone a bidirectional Glenn procedure. All operations including homograft placement in the right ventricular outflow tract, whether during repair or during right ventricular exclusion, ended in death.
Conclusion: Right ventricular exclusion with a fenestrated tricuspid valve patch combined with right atrioplasty and right ventriculoplasty and a Blalock-Taussig shunt (Starnes procedure) has provided effective palliation for neonates presenting with critical Ebstein anomaly and a tricuspid valve that cannot be repaired.