Prognostic Value of Amniotic Fluid Viral Load to Predict Adverse Outcome in Pregnancies Complicated by Congenital Cytomegalovirus Infection: A Multicenter Study

Abstract

Introduction: The aim of the study was to report the prognostic value of cytomegalovirus (CMV) viral load in the amniotic fluid (AF) in predicting the outcome of infected pregnancies. Methods: Multicenter retrospective study involving 11 Italian referral centers from 2012 to 2021 was conducted. Inclusion criteria were fetuses with confirmed congenital CMV infection. The primary outcome was the prognostic value accuracy of CMV quantitative polymerase chain reaction (qPCR) in AF in predicting the risk of additional anomalies detected either at follow-up ultrasound or fetal magnetic resonance imaging (MRI). The secondary outcome was prediction of postnatal clinical symptoms related to CMV infection. Multivariate logistic regression and area under the curve (AUC) analyses were used to analyze the data. Results: 104 fetuses were included. Associated anomalies detected at follow-up ultrasound or fetal MRI were detected in 14.4% of cases (15/104). Mean AF CMV viral load was significantly higher in fetuses with additional anomalies compared to those without additional anomalies at follow-up ultrasound or fetal MRI (3,346,634.27 ± 402,582.95 vs. 761,934 ± 222513,2 p < 0.001). At multivariate logistic regression analysis, CMV AF viral load was independently associated with the presence of additional anomalies at follow-up ultrasound or MRI, with an OR of 1.07 (p = 0.010), while maternal age (p = 0.24), trimester at maternal infection (p = 0.97), and type of infection (primary vs. non-primary) (p = 0.12) were not. CMV AF viral load had AUC of 0.755 for the occurrence of anomalies due to CMV infection, with an optimal cut-off point of >1,310,520 copies/mL, a sensitivity of 66.7%, a specificity of 84.3%, and a positive likelihood ratio of 4.24. Once excluding fetuses with anomalies at ultrasound or MRI, the diagnostic performance of qPCR in identifying fetuses with symptomatic infection after birth was low, with an AUC of 0.586. Conclusion: CMV viral load at second trimester amniocentesis has a moderate accuracy for the occurrence of CMV-related anomalies in fetuses with congenital infection and normal ultrasound at the initial diagnosis. Conversely, prediction of symptomatic infection is low.