A Prospective, Multi-Site Study of Performance of Cell-Free DNA Testing for Recessive Conditions in a Large, General-Risk Pregnancy Population.

Abstract

To evaluate the clinical performance of cell-free DNA (cfDNA) screening as a primary screening tool for autosomal recessive conditions in a large, prospective, general-risk population. We conducted a prospective analysis of pregnant individuals who underwent carrier screening with a reflex cfDNA fetal risk assessment at nine participating U.S. institutions. The conditions evaluated included cystic fibrosis, spinal muscular atrophy, and beta and alpha hemoglobinopathies. The cfDNA results were categorized as high risk (at least 1-in-4 predicted fetal risk) or low risk. Test performance metrics were calculated using Wilson Score 95% CIs. There were 2,403 cfDNA results among 2,212 pregnant carriers with singleton pregnancies at a gestational age of at least 10 weeks at participating sites (188 individuals were carriers for more than one condition). The neonatal or fetal outcome was known for 2,369 (98.6%) cfDNA results. A cfDNA fetal risk of at least 1 in 4 was returned for 1.3%, and the cfDNA fetal risk test had a specificity of 99.5%, a negative predictive value of 99.9%, a sensitivity of 94.4%, and a positive predictive value (PPV) of 58.6%. In this large, racially and ethnically diverse general-risk pregnant carrier sample, cfDNA fetal risk assessment accurately distinguished high-risk and low-risk pregnancies with greater PPV than traditional carrier screening, without requiring a partner sample. Combined with prior peer-reviewed studies that assessed cfDNA fetal risk screens, these findings support the possibility of cfDNA fetal risk assessment as a primary screen to identify pregnancies at high risk for autosomal recessive conditions.