Natural history of 107 cases of fetal aortic stenosis from a European multicenter retrospective study

Fetal aortic valvuloplasty (FV) aims to prevent fetal aortic valve stenosis progressing into hypoplastic left heart syndrome (HLHS), which results in postnatal univentricular (UV) circulation. Despite increasing numbers of FVs performed worldwide, the natural history of the disease in fetal life remains poorly defined. The primary aim of this study was to describe the natural history of fetal aortic stenosis, and a secondary aim was to test previously published criteria designed to identify cases of emerging HLHS with the potential for a biventricular (BV) outcome after FV.

From a European multicenter retrospective study of 214 fetuses with aortic stenosis (2005-2012), 107 fetuses in ongoing pregnancies that did not undergo FV were included in this study and their natural history was reported. We examined longitudinal changes in Z-scores of aortic and mitral valve and left ventricular dimensions and documented direction of flow across the foramen ovale and aortic arch, and mitral valve inflow pattern and any gestational changes. Data were used to identify fetuses satisfying the Boston criteria for emerging HLHS and estimate the proportion of these that would have been ideal FV candidates. We applied the threshold score whereby a score of 1 was assigned to fetuses for each Z-score meeting the following criteria: left ventricular length and width > 0; mitral valve diameter > -2; aortic valve diameter > -3.5; and pressure gradient across either the mitral or aortic valve > 20 mmHg. We compared the predicted circulation with known survival and final postnatal circulation (BV, UV or conversion from BV to UV).

Among the 107 ongoing pregnancies there were eight spontaneous fetal deaths and 99 livebirths. Five were lost to follow-up, five had comfort care and four had mild aortic stenosis not requiring intervention. There was intention-to-treat in these 85 newborns but five died prior to surgery, before circulation could be determined, and thus 80 underwent postnatal procedures with 44 BV, 29 UV and seven BV-to-UV circulatory outcomes. Of newborns with intention-to-treat, 69/85 (81%) survived ≥ 30 days. Survival at median 6 years was superior in cases with BV circulation (P = 0.041). Those with a postnatal UV circulation showed a trend towards smaller aortic valve diameters at first scan than did the BV cohort (P = 0.076), but aortic valve growth velocities were similar in both cohorts to term. In contrast, the mitral valve diameter was significantly smaller at first scan in those with postnatal UV outcomes (P = 0.004) and its growth velocity (P = 0.008), in common with the left ventricular inlet length (P = 0.004) and width (P = 0.002), were reduced significantly by term in fetuses with UV compared with BV outcome. Fetal data, recorded before 30 completed gestational weeks, from 70 treated neonates were evaluated to identify emerging HLHS. Forty-four had moderate or severe left ventricular depression and 38 of these had retrograde flow in the aortic arch and two had left-to-right flow at atrial level and reversed a-waves in the pulmonary veins. Thus 40 neonates met the criteria for emerging HLHS and BV circulation was documented in 13 (33%). Of these 40 cases, 12 (30%) had a threshold score of 4 or 5, of which five (42%) had BV circulation without fetal intervention.

The natural history in our cohort of fetuses with aortic stenosis and known outcomes shows that a substantial proportion of fetuses meeting the criteria for emerging HLHS, with or without favorable selection criteria for FV, had a sustained BV circulation without fetal intervention. This indicates that further work is needed to refine the selection criteria to offer appropriate therapy to fetuses with aortic stenosis. Copyright © 2016 ISUOG. Published by John Wiley & Sons Ltd.