4 DISCUSSION
In fetuses with d-TGA, severe hypoxemia may occur in the neonatal period due to inadequate blood mixing at the atrial level in the absence of VSD, making early detection and follow-up essential in experienced centers. In our study population, the FO diameter and the absence of a VSD were predictors of urgent BAS in fetuses with d-TGA.
According to definitions by Rudolph et al., 10,11 in fetuses with d-TGA with an IVS, cardiovascular circulation is characterized by blood reaching the left atrium via the FO and the left ventricle carriers oxygenated blood to the pulmonary circulation and then to the descending aorta. This condition results in restriction of the DA and increased pulmonary venous return to the left atrium, which may be associated with narrowing and restriction of the FO owing to the elevated pressure in the left atrium 12-14. In our study, the FO diameter was significantly decreased in the urgent BAS group, compared with the fetuses without urgent BAS (5.1 mm vs 6.3 mm). This finding is consistent with a report of 45 fetuses with d-TGA and IVS, in which the median FO diameter was significantly decreased (5.7 mm), requiring urgent BAS 15. Our study group also included fetuses with both VSD and IVS. In another study, the median FO diameter was 4.8 mm in the urgent BAS group as compared with 5.9 mm in fetuses not required urgent BAS, but the differences were not significant 16.
In ROC analysis, a FO diameter of 6 mm was found to predict urgent BAS with a sensitivity of 73.3% and specificity of 72.2%. In our experience, a FO diameter of greater than 6 mm highly reduced the need for urgent BAS. Similar to our finding, in a study of 60 fetuses, the FO diameter was significantly smaller in the urgent BAS group and the FO diameter was found to be the most valuable predictor for urgent BAS, though their cut off value on the ROC analysis was greater (6.5 mm). This difference may arise from the incidences of VSD, being 41.7% in the current study vs 18% in their study group17. Besides FO, some authors also used the ratio of FO to the total septal length (FO:TCL), with a FO:TCL of 0.5 yielding a higher predictive value for urgent BAS with a sensitivity of 99%, specificity of 60%, positive predictive value of 50% and negative predictive value of 94%18.
In our study, 5 fetuses had restrictive appearance of FO and no high-flow rate in the DA, suggestive of absence of constriction of the DA. While one fetus had an aneurysmatic FO with its flap extending more than 50% to the left atrium, the remaining four had a flat FO flap swinging less than 30° into the left atrium. In accordance with other studies, all 5 fetuses underwent urgent BAS immediately after birth16,18. There are also some studies suggesting that the appearance of the FO flap (flat, fixed or aneurysmatic) was not associated with urgent BAS 19-22. Inconsistencies between studies may result from the small number of fetuses included, the definition of urgent BAS, the lack of echocardiographic examination after 37 weeks of gestation, or misdiagnosis of FO restriction in normal fetuses due to breathing movements and physiological right ventricular hypertropia. In addition, interpretation of the FO may vary due to the primary or secondary nature of restriction: the former represents an abnormally narrowed aneurysmatic FO and motility, while the latter may result from hemodynamic alterations including abnormal blood flow in the DA and pulmonary venous return 23. In our study, in 13 fetuses requiring urgent BAS, narrowing of the FO was possibly caused by underlying d-TGA hemodynamics, i.e., the pumping of blood with high oxygen saturation directly from the left ventricle to the lungs, causing increased pulmonary venous return to the left atrium.
In our experience, the presence of a VSD appeared to play a protective role from urgent BAS: only 4 of 20 fetuses with VSD underwent urgent BAS. In a retrospective single center study, BAS was performed in 64% of neonates with an IVS and 25% of neonates with VSD24. In addition to the FO diameter, the absence of VSD was found to be an independent predictor of urgent BAS (Table 7). In the ROC analysis, the cut off value for the VSD diameter was 3.2 mm, with 75% sensitivity and 75% specificity.
BAS can generally be thought of as a safe procedure. In our study, no complications developed in the neonates who underwent urgent BAS. In a study of 73 neonates undergoing BAS, the procedural success rate was 98.6%, while hemodynamically significant arrhythmia developed in 4.1%, tamponade in 1.4%, and catheter-related complications in 2.7%24.
The strengths of this study are the comprehensive analysis of the FO and VSDs with echocardiographic examinations performed most commonly after 37 weeks of gestation and a single center study. There are two limitations to this study. First, its retrospective design. Second, we could not include the FO in ROC and regression analyses in fetuses with VSD or IVS separately due to the small number of the fetuses.
In conclusion, prenatal echocardiography performed after 37 weeks of gestation in fetuses with d-TGA provides valuable information about the dimensions of FO and the absence of VSD to estimate the need for postnatal urgent BAS that would prevent immediate life-threatening complications.