Echocardiography. 2025 Oct;42(10):e70322. doi: 10.1111/echo.70322.
ABSTRACT
OBJECTIVE: To evaluate whether combining measurements of ventricular and atrial size and shape obtained from the four-chamber view using speckle tracking analysis improves the prediction of which fetuses with D-transposition of the great arteries (D-TGA) will require urgent neonatal balloon atrial septostomy (BAS).
METHODS: A retrospective study was conducted using fetal echocardiography databases from two tertiary care centers. Fetuses with D-TGA and an intact ventricular septum were included; those with additional cardiac anomalies were excluded. Subjects were categorized into two groups based on postnatal outcomes: those requiring urgent BAS within 24 h of birth and those who did not. Digital four-chamber view images from the last prenatal examination were analyzed using speckle tracking software to measure ventricular and atrial end-diastolic size and shape, including chamber length, width, area, and sphericity index (SI). Z-scores were calculated from normative data derived from 200 control fetuses. Logistic regression was used to identify significant predictors of BAS, and a probability model was created using the regression coefficients. Receiver operating characteristic (ROC) curve analysis was used to determine optimal probability thresholds, and the Kolmogorov-Smirnov test compared z-score distributions between groups.
RESULTS: Thirty-nine fetuses with D-TGA were included; 22 (56%) underwent urgent BAS. Logistic regression identified specific ventricular and atrial parameters, particularly the left ventricular mid-chamber sphericity index and right atrial sphericity index, as significant predictors. A probability threshold of ≥65% predicted BAS with 96% sensitivity (95% CI: 77%-99%) and 88% specificity (95% CI: 67%-98%), with an overall accuracy of 92% and area under the ROC curve of 0.93 (p < 0.0001). The Kolmogorov-Smirnov test confirmed significant shape differences between BAS and non-BAS groups, with BAS fetuses showing flatter left ventricles and more globular right atria. The analysis was performed using standard grayscale ultrasound images, and the required measurements can be obtained with routine imaging tools available in most obstetric practices. An Excel-based calculator was developed to facilitate clinical use.
CONCLUSION: Combining atrial and ventricular end-diastolic length and area measurements significantly improves the prediction of which fetuses with D-TGA will require urgent postnatal BAS. This method offers higher sensitivity and specificity than previously published approaches. Because it relies only on basic four-chamber imaging, it is feasible for widespread clinical application and can aid in timely perinatal management planning for affected fetuses.
PMID:41105322 | DOI:10.1111/echo.70322
