A prospective evaluation of exome sequencing in 51 fetuses with multiple congenital anomalies and lessons for future prenatal implementation.

Abstract

Introduction: Exome sequencing (ES) is a powerful adjunct to post-mortem examination for investigating fetuses with congenital anomalies. However, with the implementation of rapid "in-utero" sequencing becoming eminently feasible, fetal phenotyping may, in future, rely on antenatal imaging alone. Aim(s): To determine whether the diagnostic yield of ES in fetuses with congenital anomalies following autopsy is comparable to the yield of ES implemented at the time of antenatal ultrasound detection. Method(s): ES was prospectively performed on 51 fetal probands following autopsy as part of the Perinatal Autopsy Flagship of the Melbourne Genomics Health Alliance Demonstration Project. Candidate variants concordant with phenotypic findings from fetal autopsy were classified by multidisciplinary review using ACMG guidelines. Independently, a phenotype-driven virtual gene panel for each proband was derived from antenatal imaging reports by clinicians blinded to autopsy findings. Where a pathogenic variant fell within this "pre-autopsy" virtual gene panel, this was viewed as a diagnosis that would have been made by combining WES with antenatal imaging alone. Result(s): The diagnostic rate in this cohort was 17/51 (33%) with 4 further variants awaiting functional validation. Phenotypes with the highest diagnostic yield included skeletal dysplasias (100%); multiple malformations (33%) and hydrops fetalis (33%). 14/17 of diagnoses would have been made on antenatal findings alone while 3/17 additional diagnoses were made with findings only apparent on autopsy. Conclusion(s): ES in pregnancy following the detection of structural anomalies on ultrasound is feasible and returns comparable diagnostic yields to ES combined with autopsy.