Prenatal exome sequencing and impact on perinatal outcome: cohort study

ABSTRACT Objectives First, to determine the uptake of prenatal exome sequencing (pES) and the diagnostic yield of pathogenic (causative) variants in a UK tertiary fetal medicine unit following the introduction of the NHS England Rapid Exome Sequencing Service for fetal anomalies testing (R21 pathway). Second, to identify how the decision to proceed with pES and identification of a causative variant affect perinatal outcomes, specifically late termination of pregnancy (TOP) at or beyond 22 weeks' gestation. Methods This was a retrospective cohort study of anomalous fetuses referred to the Liverpool Women's Hospital Fetal Medicine Unit between 1 March 2021 and 28 February 2022. pES was performed as part of the R21 pathway. Trio exome sequencing was performed using an Illumina next‐generation sequencing platform assessing coding and splice regions of a panel of 974 prenatally relevant genes and 231 expert reviewed genes. Data on demographics, phenotype, pES result and perinatal outcome were extracted and compared. Descriptive statistics and the χ‐square or Fisher's exact test were performed using IBM SPSS version 28.0.1.0. Results In total, 72 cases were identified and two‐thirds of eligible women (n = 48) consented to trio pES. pES was not feasible in one case owing to a low DNA yield and, therefore, was performed in 47 cases. In one‐third of cases (n = 24), pES was not proposed or agreed. In 58.3% (14/24) of these cases, this was because invasive testing was declined and, in 41.7% (10/24) of cases, women opted for testing and underwent chromosomal microarray analysis only. The diagnostic yield of pES was 23.4% (11/47). There was no overall difference in the proportion of women who decided to have late TOP in the group in which pES was agreed compared with the group in which pES was not proposed or agreed (25.0% (12/48) vs 25.0% (6/24); P = 1.0). However, the decision to have late TOP was significantly more frequent when a causative variant was detected compared with when pES was uninformative (63.6% (7/11) vs 13.9% (5/36); P