Can Prenatal Diagnosis of Total Anomalous Pulmonary Venous Return (TAPVR) Using Routine Fetal Ultrasound be Improved? A Case-Control Study

Abstract

Objectives. To determine the most common fetal ultrasound markers of total anomalous pulmonary venous return (TAPVR) during mid-trimester ultrasound using standardly obtained images and evaluate the performance of diagnostic algorithms for improving prenatal diagnosis. Methods. This was a matched case-control study at a regional referral centre (2005 to 2019). Cases of TAPVR were matched to controls 1 : 4 by date of birth and biologic sex. Postprocessing review of stored fetal ultrasound images was performed by two blinded and independent observers in a standardized fashion using nine sonographic markers: (i) left/right heart disproportion; (ii) abnormal distribution of great vessels; (iii) pulmonary vein entry into the left atrium (LA); (iv) confluence behind the LA; (v) abnormal coronary sinus; (vi) absence of the Coumadin ridge; (vii) aortic diameter; (viii) distance between LA and aorta; and (ix) post-LA space index >1.27. Descriptive and inferential statistics were used to present results and compare cases and controls. Diagnostic algorithms were compared by sensitivity/specificity. Results. 21 cases of isolated TAPVR were matched to 84 controls (n = 105). The most common ultrasound marker of TAPVR was absence of pulmonary vein entry into the LA (42.9%), followed by abnormal Coumadin ridge (38.1%). Cases of TAPVR had significantly larger post-LA spaces than controls ( $p<0.0001$ ) and wider aortic diameters ( $p=0.006$ ). A diagnostic algorithm stratifying on absence of pulmonary veins followed by an abnormal Coumadin ridge, can correctly identify cases of TAPVR with high specificity (90.5%) and moderate sensitivity (61.9%). Conversely, a diagnostic algorithm using the presence of any 3 abnormal markers had improved specificity (94.1%) but poorer sensitivity (23.8%). Conclusions. Using standardly obtained images from routine fetal ultrasound, improved prenatal detection of isolated TAPVR is possible. A standardized diagnostic approach can be highly specific for fetal TAPVR, however, algorithms that are sufficiently sensitive for screening in the general population are still needed.