Propensity Score Matched Analysis of Cleft Closure in Complete Atrioventricular Septal Defect Repair

Edward Buratto, MBBS, PhD, Adrienne Lui, MD, Thomas Hu, MD, Phillip S. Naimo, MD, PhD, Yaroslav Ivanov, MD, PhD, Yves d’Udekem, MD, PhD, Christian P. Brizard, MD, and Igor E. Konstantinov, MD, PhD


Take Home Points:

  • Reoperation on left atrioventricular valve (LAVV) is not negligible after repair of complete atrioventricular septal defect (cAVSD).
  • Whether or not to close the LAVV cleft does not impact on need for reoperation.

Dr. Yasuhiro Kotani

Commentary from Dr. Yasuhiro Kotani (Okayama, Japan), chief section editor of Congenital Heart Surgery Journal Watch:



Single center retrospective study included 455 patients who underwent cAVSD repair between 1990 and 2019. To determine the effect of cleft closure, propensity score matching was performed on risk factors for reoperation after cAVSD repair. Median age was 3.6 months (mean, 9.6 ± 20.4 months) and median weight was 4.3 kg (mean, 4.7 ± 4.3 kg). Early mortality was 2.9% (13 of 455), and survival was 89.8% ± 1.9% at 20 years. Early reoperation was a risk factor for mortality (P=.004). Freedom from LAVV reoperation was 74.1% ± 4.0% at 20 years. Preoperative severe LAVV regurgitation (P<.001) and early postoperative moderate or greater LAVV regurgitation (P=.007) were risk factors for reoperation. Propensity score matching included 106 pairs. There were no differences in long-term survival (P=.71) or reoperation (P=.26) between the 2 groups.



This retrospective study was done at Royal Children’s Hospital, Melbourne. Outcomes were acceptable, including about 75% of patients remain free from LAVV reoperation 20 years after initial cAVSD repair. The study included the propensity score matching to see the impact of the closure of the LAVV, which showed no difference in need of reoperation. There was a significantly higher rate of preoperative moderate or greater LAVVR in patients who had a cleft closure (25.6% vs. 15.9%, p=.03). In contrast, patients without cleft closure had more deficient lateral leaflet (7.5% vs. 3.2%, p=.50), which infers those with significant LAVVR had smaller lateral leaflet as etiology of regurgitation.


The results in this study raised 2 questions. Firstly, it is unclear that cleft closure was cause or effect. We can speculate that cleft closure was performed in patients with preoperative significant LAVVR as most of surgeons would not leave the cleft unrepaired. Therefore, we could think that cleft closure is the result. Secondary, 2 groups were unbalanced in terms of era of operation (p<.001), but there were no differences in survival and freedom from LAVV reoperation. This study showed that recent era of surgery was protective in terms of risk for reoperation. There were more patients with cleft closed in the most recent era. These facts make us think that surgical strategy to close the cleft may influence the outcome in this study. Taken together, prospective study to see the effect of LAVV closure is necessary. Clinically, it is challenging to repair the LAVV when patients have significant LAVVR with small lateral leaflet, as closure of the cleft may make LAVVR worse.