Dr. Kaiser, Dr. Pairolero, members and guests:Surgical repair of primum atrial septal defects usually is simple and yields good results. These results continue to improve in the current era. Certain sub-groups of children carry higher mortality, in particular, those children who present in the first year of life with congestive heart failure. Late follow-up of children after repair of primum atrial septal defect indicate a definite incidence of reoperations. The purpose of our study was to examine trends in early mortality in the current era of better understanding of left atrioventricular valve morphology and better myocardial protection as well to elucidate whether the previously reported risk factors continue to predict outcome.
From 1982 until 1996, 180 children with Primum atrial septal defect and a partitioned common A-V valve were included in the study. Children with inlet inter-ventricular communication, unbalanced defects leading to uni-ventricular repair, A-V & V-A discordance, and children with total anomalous pulmonary venous connection were excluded. These children were excluded because the pathophysiology of these lesions would determine the outcome of repair.
This was a retrospective review. 30 variables were analyzed and examined to determine predictors of death, re-operation, and left A-V valve regurgitation.
The mean age of repair was 4.6, ranging from 1 month to 16 years. 36% were male and 25% had Down syndrome. At presentation, absent or mild symptoms were present in 80% and, in 20%, severe symptoms or congestive heart failure was present.
This is a graph of age distribution, on the X-axis are the age groups. And Y-axis are the number of children. As you can see, most children were repaired under the age of five years. We had 22 children who were repaired under the age of one year.
Preoperative cardiac catheterization was performed on 60 patients and this has shown a preoperative mean pulmonary artery pressure of 20. Preoperative left A-V valve regurgitation as graded by echocardiography revealed absent or mild regurgitation in 63%, moderate in 2%, and severe in 2%.
A whole array of associated anomalies was present. These are the some of the important ones: double orifice left A-V valve was present in 5%; sub-aortic obstruction was present in 4% at presentation; and borderline hypoplastic ventricle, and by that I mean between 70%-90% predicted, was present in 2.6%; and two patients had a parachute left A-V valve.
During repair, autologous pericardium was used for patching in 95% and the other 5% had Dacron patch closure. We imply an individualized approach to the management of left A-V valve. We indeed believe in Dr. Carpentier's principle in that the left A-V valve is a tri-leaflet valve, rather than a mitral valve with a cleft. So the extent of cleft closure is determined by multiple factors of which the most important are the degree of pre-operative regurgitation, the morphology of the valve, and associated anomalies such as parachute left A-V valve or double orifice left A-V valve. In this series, we had 59% who had complete cleft closure, 29% had partial, and 12% had no closure of the cleft. We had three patients who had normally developed mitral and tricuspid valves with no clefts. This entity has been previously reported by Professor Anderson and may represent a form frust type primum atrial septal defect.
Associated procedures were performed on 25 patients. 7 had closure of muscular VSD's and 6 had concomitant subaortic resection at the time of initial repair. And in patients with a borderline ventricle an injunctive bi-directional cavopulmonary shunt or patch fenestration was performed.
Three patients died early after the operation, accounting for a mortality of 1.6%. Two of these patients were in the infant group and died early in the series, before 1986. Atrial arrhythmias occurred in 3% and transient A-V block occurred in 1.6% percent. None of these patients required permanent pacemaker insertion early after the operation. Nevertheless, two patients had a pacemaker insertion; one at the time of repair for a pre-existing Sick sinus syndrome. And the other was nine years after the operation for late development of A-V block.
The follow up was up to 14.4 years with a mean of 6 years. Echocardiographic follow up was available in 80% of children at a mean of 4.6 years. Left A-V valve regurgitation was graded as absent or mild in 85%, moderate in 14% and severe in 1%. These two patients would be put forward very soon for re-operation. These results are statistically different from the pre-operative data.
The overall survival at 10 years was 98% including operative mortality.
Nineteen children underwent re-operations, 15 of them were for left A-V valve regurgitation and 6 for sub-aortic stenosis.
The mean interval from the initial repair to re-operation for left A-V valve regurgitation was 2.4 years. The findings at the re-operation were residual cleft in 9, disrupted cleft sutures in 2, obvious annular dilatation in 3. 12 of these patients were repaired and one had a left A-V valve replacement.
Worth noting, that there was only one child who had isolated cleft closure at re-operation. The other 11, in addition to cleft closure had other reparative techniques such as annuloplasty, commissuroplasty, or chordal shortening. This suggests to us that although residual clefts has been reported to be a predictor for late left A-V valve regurgitation, nevertheless, annular dilatation and chordal elongation may play an important roll.
Freedom from re-operation for left A-V valve regurgitation at 10 years was 89%.
Six children underwent re-operation for subaortic stenosis at a mean interval of 5.7 years. All of them had subaortic resection and, in one patient, in addition to sub-aortic resection, had an infundibular patch or what is called a modified Konno procedure with preservation of the aortic valve.
Freedom from re-operation for sub-aortic stenosis at 10 years was 95%, and at 14 years was 90%.
Because of the reported high mortality of the infant group, which reaches up to 35% in some of the large series, we looked at this specific sub-group and compared them with other children. We listed only the statistically significant variables. The infants had higher incidence of congestive heart failure and higher incidence of borderline hypoplastic ventricles. They had higher pulmonary to aortic pressure ratio, higher incidence of death, and re-operation.
In the univariate analysis for predictors of death, severe symptoms approached statistical significance and age below one year was statistically significant. In the multi-variate analysis age below one year was the only predictor of death. In the uni- and mutli-variate analyses for re-operation for left A-V valve, age below one year and pre-operative moderate to severe left A-V valve regurgitation were statistically significant.
In conclusion, Repair of primum atrial septal defect can be accomplished with low mortality. Re-operation for left A-V valve regurgitation and sub-aortic obstruction are important complications. However they can be repaired safely with no mortality. Presentation the first year of life carries a worse outcome, although the mortality has improved in recent years.
I would like to thank the Society for the privilege of presenting our data.