Introduction

Atrioventricular septal defect (AVSD) is a defect of the common atrioventricular junction with deficient atrioventricular septation. It accounts for 3% of all heart defects, with a strong coindication with Down syndrome (DS), occurring in 30%–40% of children with trisomy 21.1 These patients account for 70% of individuals with AVSD.2 Risk factors for nonsyndromic AVSD appear to be offspring and maternal pregestational diabetes, gestational diabetes, and obesity.3 The key element of complete AVSD is the presence of a single atrioventricular valve common to both the right and left chamber.

Patients undergo total surgical correction, usually with a double-patch or, less commonly, a single-patch technique. Results of surgical repair for complete AVSD have improved markedly in the past 30 years,4 and current mortality rate has dropped to 2.5%.5 Some studies suggest different early and late outcomes of surgery in children with or without trisomy 21.6 A significant number of patients after correction suffer from mitral valve (MV) dysfunction, and moderate to severe left atrioventricular valve regurgitation occurs in 21% of cases.7 It is also the most common reason for reoperation, usually years after the primary surgery.8 Less often, patients require pacemaker implantation and relief of left ventricular outflow tract obstruction (LVOTO). In certain cases, ventricular imbalance occurs, and surgical correction must be performed in the direction of the Fontan tract.

The aim of the study was to evaluate late outcomes of surgical correction of AVSD in patients with and without DS, particularly the frequency of reoperation due to mitral dysfunction, and late rhythm disturbances requiring pacemaker implantation over a total of 26 years.

Patients and methods

Our retrospective analysis included a total of 86 consecutive patients with complete AVSD undergoing corrective surgery between 1996 and 2022 in a tertiary congenital heart defects center. Patients with an unbalanced form of complete AVSD, partial type of the defect, and patients with other significant heart diseases were excluded from the study. The analyzed group comprised 48 boys and men and 38 girls and women at a median age of 10.3 (interquartile range [IQR], 1.9–27.5) years. Median age at the time of the surgery was 5 (IQR, 1–126.6) months. In the whole group, 68.6% (n = 59; 33 boys and 26 girls) of patients were diagnosed with DS, and the remaining 31.4% (n = 27; 14 boys and 13 girls), had a normal karyotype. All the patients underwent double-patch surgery, the atrial septal defect (ASD) was usually closed with autogenic pericardium and the ventricular septal defect (VSD) with a Dacron patch. Diagnostic workup for all individuals included electrocardiogram, chest X-ray, and full echocardiographic examination with detailed evaluation of the common atrial-ventricular valve. The last available transthoracic echocardiography examination was analyzed for postoperative detailed MV long-term function. Regurgitation was defined as mild when the central regurgitation jet was limited to the perivalvular area, moderate when the regurgitation jet extended to half the length of the left atrium, and severe when the huge central regurgitation jet reached the posterior wall of the left atrium. Median follow-up time was 8.5 (IQR, 1–27) years; 10 (IQR, 1–27) years in the DS group and 8 (IQR, 2–22) years in the patients with a normal karyotype. The research project was approved by the Medical University of Silesia in Katowice (KNW/0022/KB1/95/14). Parents or patients were informed, and anonymous retrospective data were gathered.

Statistical analysis

Differences between the groups were tested for significance with the Fisher exact test. Due to a small sample size, we decided not to perform a correction for multiple testing. The Kaplan–Meier freedom from end point was evaluated with the log-rank test for curve comparison. The end point was defined as the first major reoperation at follow-up: redo MV, LVOT relief, or residual VSD closure. The statistical analyses were performed using MedCalc Statistical software version 20.218 (MedCalc Software Ltd, Ostend, Belgium). P values below 0.05 were considered significant.

Results

All 86 patients underwent total biventricular correction, with the defect operated using the double-patch technique. Pulmonary artery banding was used as a staged approach to complete repair in 9 patients (10.5%), with a similar frequency in the children with and without trisomy 21 (10% and 11.1%, respectively; P >0.99). Postoperative mitral regurgitation (MR) is a major late surgical adverse effect that not only significantly impacts patients’ clinical status but also represents the most common cause for reoperation. MV regurgitation was detected in all patients on echocardiography conducted at their last follow-up visit.

Among the patients with trisomy 21, MR was graded as mild in 25 individuals (42.37%), moderate in 26 (44.07%), and severe in 8 cases (13.56%). Among the patients with a normal karyotype, 10 (37.04%) had mild MR, 11 (40.74%) moderate, and 6 (22.22%) severe MR. There were no significant differences between the groups.

Eleven patients had a residual hemodynamically insignificant ventricular septal defect, 6 in the DS group and 5 with a normal karyotype. Five patients had LVOTO, and surgical relief was performed in 2 of them (1 with and 1 without DS). For the other 3 patients, the gradient levels were below the threshold required for surgical intervention. Within the entire group, 11 patients (12.8%) needed reoperations, which involved either mitral valvuloplasty or valve replacement, and relief of LVOTO. They were performed more frequently in the patients with a normal karyotype (25.9% vs 6.8%; P = 0.03).

Altogether, 9 patients underwent MV repairs after complete correction, more often the non-DS individuals (22.2% vs 5.1%; P = 0.02). Two patients underwent subsequent reoperation of MV replacement, and both had a normal karyotype. Median time to MV reoperation from primary repair was 20 (IQR, 3–170) months in the patients with DS and 17 (IQR, 1–42) months in the patients with a normal karyotype. During 15-year follow-up, 82.2% of the patients did not need a reintervention. Among those, the patients with trisomy 21 showed a notably higher rate of free from reintervention cumulative survival at 87.5%, as compared with 70.7% in the patients without trisomy 21 (P = 0.01). Hazard ratio (HR) for any reintervention in the patients with trisomy 21 was 0.18 (95% CI, 0.05–0.67). A rate of cumulative survival free from MV reoperation reached 85.3%, with a considerable difference between patients with and without DS (74.9% vs 90.2%, respectively; P = 0.02). Accordingly, HR for MV reintervention in the patients with trisomy 21 was 0.17 (95% CI, 0.04–0.71). The Kaplan–Meier curves are presented in Figure 1. Due to conduction disorders, pacemakers were implanted in 11.63% (n = 10) of patients. The device was implanted in 9 DS patients, and 1 patient with a normal karyotype (15.25% vs 3.7%; P = 0.16). In most patients, the pacemaker implantation should be treated as a mid-term complication, with median time between the surgical correction and pacemaker implantation reaching 46 (IQR, 1–228) months. One patient with a normal karyotype received a pacemaker after 16.3 years, which can be classified as a late complication.

Figure 1. Freedom from reintervention and mitral valve (MV) reoperation in patients with and without trisomy 21. Kaplan–Meier curves with 95% CI

Discussion

Advances in the surgical treatment of congenital heart defects resulted in an improved, consistently low mortality rate.9 This study reflects an 18-year long, single-center experience of AVSD repair using the double-patch technique. While some centers may prefer a single-patch repair, studies show no significant difference in mortality, residual shunts, postsurgical valve regurgitation, LVOTO, or heart block between the 2 techniques,10,11 suggesting that these findings are applicable to the general population. Numerous sources report a better prognosis for patients with AVSD and coexisting DS.12-14 Authors mostly indicate a lower mortality rate, but the question of reoperation incidence is debatable depending on the source.15 Quite a few papers include a comparison of reoperation frequency in patients with and without trisomy 21. Our research clearly highlights a disparity in reoperation likelihood between the patients with and without trisomy 21, drawing particular attention to significantly higher occurrence of MV repairs in children without DS. This aligns with surgeons’ observations that heart valve tissue in children with DS tends to be slightly thicker and “augmented,” as compared with that of children with a normal karyotype, facilitating easier MV reconstruction. Despite this, MV replacement was infrequently necessary, required only in 2 patients with a normal karyotype who previously underwent MV valvuloplasty; none DS patient needed this procedure. Some authors point out to differences in histologic composition and collagen structure abnormalities in DS patients, which may contribute to better valve coaptation and, consequently, less mitral regurgitation,13 potentially affecting valve replacement likelihood in patients with a normal karyotype. The need for reoperation depends not only on karyotype status. Risk factors include incomplete cleft closure, the presence of a double orifice left atrioventricular valve, and other associated cardiovascular anomalies.6,16 Our study aimed for a homogenous group, excluding patients with other heart anomalies requiring correction and those with a double orifice MV. Despite this, varying outcomes of corrective surgery have been reported. Stulak et al8 observed a significantly higher proportion of DS patients undergoing both valvuloplasty and valve replacement, which contrasts with our findings. This discrepancy could be due to factors other than trisomy 21 and additional heart diseases. Suzuki et al17 highlighted preoperative regurgitation as a risk factor for postoperative MV regurgitation, while Al-Haay et al18 noted higher regurgitation occurrence due to more common MV dysplasia in patients without trisomy. The higher number of MV reoperations in patients with a normal karyotype, may be related to higher incidence of MV dysplasia in these patients or higher preoperative incidence of regurgitation, which may also be due to different anatomy of the defect. Ten Harkel et al14 associated a superior freedom from operation in DS patients with a greater number of preoperative MV regurgitations, suggesting that leaflet extension and mobility properties are unfavorable in the native common valve. However, the same properties may actually result in better funcioning of the repaired valve. This is supported by our findings, where DS patients showed higher incidence of mild MV regurgitations than those with a normal karyotype (42.4% vs 30%). Nevertheless, as a potential late complication, the number of reoperations is expected to rise with extended follow-up. Yet, similarly to other studies,19 our research showed a freedom from reoperation of over 82% at 15 years, with the value higher in DS patients (87.5% vs 70.7%; P = 0.01). A reverse situation was observed for the conduction disorders. Pacemakers were implanted in 9 DS patients and in only 1 patient with a normal karyotype (15.25% vs 3.7%; P = 0.123). This may result from different defect anatomy and more extensive surgery. Moreover, patients with DS have been described to have various abnormalities regarding the amount and structure of collagen.20 With extra chromosome 21, increased expression of collagen VI in the patients with DS may be responsible for fibrosis and heart block. This may require further anatomic and histopathologic studies.

Study limitations

The study’s limitations include the genetic burden and profound disabilities associated with DS and variable length of follow-up, which might underrepresent the number of patients needing MV repair. It is also important to acknowledge a significantly smaller number of patients without trisomy 21, as compared with the DS group, although this distribution mirrors that in the affected population.

Conclusions

Patients with a normal karyotype who underwent total biventricular AVSD correction significantly more often needed reoperations than those with DS. Additionally, a subset of these patients needed MV replacement following unsuccessful valvuloplasty reoperations. Trisomy 21 was identified as a risk factor for the need of pacemaker implantation.