It has been generally reported that the outcome of the Norwood operation with an RV-PA shunt for HLHS may be better than the Norwood with modified Blalock-Taussig shunt [4, 6, 7]; therefore, we performed the primary Norwood operation with an RV-PA shunt instead of a Blalock-Taussig shunt. However, it is also reported that the primary Norwood operation requires CPB at the neonate stage, which might lead to long-term neurodevelopmental dysfunctions [8]. In addition, the patients are in a status of pulmonary high flow, which leads to heart failure until the BDG is performed, and the subsequent failure of BDG or Fontan surgery [9].
In 2004, we changed our strategy to perform bPAB soon after birth, and the Norwood with BDG about three months after. However, after changing our strategy, we performed PTPA after Norwood with BDG for almost all patients because of the possible occurrence of pulmonary artery stenosis at the debanding site. Additionally, the venous pressure of the SVC at Norwood with BDG was high, which might lead to a high risk of stroke or BDG takedown.
In 2012, we changed our strategy to perform bilateral PA banding soon after birth, then Norwood with an RV-PA shunt about one month later, and the BDG at about 6 months of age. Using this strategy, we considered that the risk of immature of pulmonary artery will be reduced since this strategy could shorten the interval of bilateral PA banding and supply the pulsatile flow from the RV-PA shunt to reduce the risk of immature of pulmonary artery.
In order to assess growth of the pulmonary artery, we focused on the diameter of the bilateral pulmonary artery, PA index, the ratio of PTPA after BDG, and the trend of the venous pressure of SVC. We found no difference in the diameter of the bilateral pulmonary artery and the PA index between groups S and G. However, the ratio of PTPA after BDG was lower in group S than G, and the venous pressure of SVC at the time of BDG was lower in group S than G. These findings might support our assumption that shortening the bPAB period and supplying the pulsatile flow from the RV-PA shunt could bring up the pulmonary artery and reduce the influence of th banding site.
In order to assess RV function, we focused on the RVEDV, EF, CI, RVEDP, TR, and B-type natriuretic peptide (BNP). Because the BDG leads to volume reduction for the RV, the timing of the BDG should be earlier to prevent RV volume overload. In this study, we showed that the timing of BDG in group S was later than in group G, therefore, the RVEDV was higher in group S than G peri TCPC and one year after TCPC. We revealed that there were no differences in RVEF, CI, RVEDP, and BNP between group S and G, which indicated that there was little influence on the RV function at one year after TCPC; however, we need to consider that the myocardial function could be affected.
In group S, the patients were younger at the time of TCPC and the interval from BDG and TCPC was shorter than in group G; therefore, the group S patients were in the condition of low saturation for a shorter period than group G. Consequently, the hepatic factor could be supplied to the lung earlier in group S than in group G, which could reduce the risk of formation of collateral vessels and pulmonary arteriovenous fistulas. The fact that RVEDP after TCPC and cardiac return at the TCPC tended to be lower in group S than in group G supports our prediction. The duration of pleural drainage time was shorter in group S than G, with a lower incidence of chylothorax in group S than G. This indicates that we were able to perform the TCPC in good condition in group Scompared to group G
Limitations
This study has several limitations. First, it was a retrospective review of HLHS in a single center. Second, our small sample size limited our ability to perform statistical adjustments. Third, in this study, all patients underwent bPAB at first palliation. The optimal operative approach for performance of a primary Norwood in the neonatal period or perform the bilateral PAB before the Norwood remains unsettled.