Amniotic fluid microRNA profiles in twin-twin transfusion syndrome with and without severe recipient cardiomyopathy

Abstract

Twin-twin transfusion syndrome presents many challenges for clinicians, and the optimal means of identifying pregnancies that will benefit most from intervention is controversial. There is currently no clinically available biomarker to detect twin-twin transfusion syndrome or to stratify cases based on the risk factors. microRNAs are small RNAs that regulate gene expression and are biomarkers for various disease processes, including adult and pediatric heart failure. To date, no studies have investigated amniotic fluid microRNAs as biomarkers for disease severity, specifically for severe recipient cardiomyopathy in twin-twin transfusion syndrome cases. This study aimed to assess whether amniotic fluid microRNAs could be useful as biomarkers to identify pregnancies at greatest risk for severe recipient cardiomyopathy associated with twin-twin transfusion syndrome. Amniotic fluid was collected at the time of amnioreduction or selective fetoscopic laser photocoagulation from monochorionic diamniotic twin pregnancies with twin-twin transfusion syndrome at any stage. Fetal echocardiography was performed on all twins before the procedure, and severe cardiomyopathy was defined as a right ventricular myocardial performance index of the recipient fetus of >4 Z-scores. microRNA was extracted from the amniotic fluid samples and analyzed using an array panel assessing 379 microRNAs (TaqMan Open Array, ThermoFisher). Student t tests were performed to determine significant differences in microRNA expression between pregnancies with severe recipient cardiomyopathy and those with preserved cardiac function. A stringent q value of <.0025 was used to determine differential microRNA expression. Random forest plots identified the top 3 microRNAs that separated the 2 groups, and hierarchical cluster analysis was used to determine if these microRNAs properly segregated the samples according to their clinical groups. A total of 14 amniotic fluid samples from pregnancies with twin-twin transfusion syndrome with severe cardiomyopathy were compared with samples from 12 twin-twin transfusion syndrome control cases with preserved cardiac function. A total of 110 microRNAs were identified in the amniotic fluid samples. Twenty microRNAs were differentially expressed, and the top 3 differentiating microRNAs were hsa-miR-200c-3p, hsa-miR-17-5p, and hsa-miR-539-5p. Hierarchical cluster analysis based on these top 3 microRNAs showed a strong ability to differentiate severe cardiomyopathy cases from controls. The top 3 microRNAs were used to investigate the sensitivity and specificity of these microRNAs to differentiate between the 2 groups with a receiver operating characteristic curve demonstrating sensitivity and specificity of 80.8%. All 20 differentially expressed microRNAs were down-regulated in the group with severe cardiomyopathy. Amniotic fluid microRNAs demonstrated differential expression between twin-twin transfusion syndrome recipient fetuses with severe cardiomyopathy and those without and have the potential to be important biomarkers of disease severity in this population.