Abstract / Summary
Background and Objectives: Congenital cytomegalovirus (CMV) infection following primary maternal infection in the first trimester carries the highest risk of long-term sequelae, yet prenatal diagnosis depends on amniotic fluid polymerase chain reaction (PCR), which cannot be performed before 17 weeks. Detection of CMV in chorionic villi at 11–14 weeks has been proposed to bring this information forward and to exclude CMV-related embryopathy. We systematically reviewed the published studies that have evaluated this test in order to determine its diagnostic accuracy. Materials and Methods: We conducted a systematic review of diagnostic test accuracy (PROSPERO CRD420261482510), and reported according to PRISMA-DTA. Five databases were searched up to 19 August 2026 without language or date restrictions. Eligible studies enrolled pregnancies with documented primary maternal infection and allowed reconstruction of a two-by-two table against CMV DNA in neonatal urine and/or saliva within three weeks of birth (with fetal or placental tissue accepted for pregnancies ending in termination). Screening, extraction and QUADAS-2 assessment were performed in duplicate and blinded. Overlapping cohorts were reduced to the largest, most completely verified report. Wilson score confidence intervals (CIs) were used; pooling was pre-specified only for four or more independent datasets. Results: Of 704 unique records, three independent cohorts in three primary publications were eligible, contributing 340 villus samples. The largest cohort (n = 330) gave a sensitivity of 40.5% (95% CI 27.0–55.5), specificity of 99.3% (97.5–99.8), positive predictive value of 89.5%, and negative predictive value of 92.0% at a prevalence of 12.7%. The remaining cohorts contributed four and six samples in turn, all true negatives. No study was at low risk of bias in all four QUADAS-2 domains; the cohort carrying the sensitivity estimate was at high risk in two, including verification of terminated pregnancies on placental tissue, whose net effect on sensitivity cannot be determined from the published data. Certainty was very low throughout. Conclusions: CMV PCR on chorionic villi is highly specific but misses approximately three infected fetuses in five, so a negative result does not reliably exclude congenital infection, and the proposition that it excludes embryopathy is unevaluated. Based on current evidence, it should not be used as a standalone basis for first-trimester decisions about continuing a pregnancy.