Prenatal maternal plasma DNA screening for cystic fibrosis: A computer modelling study of screening performance [version 1; referees: 2 approved]

• Main Authors: Robert W. Old, Jonathan P. Bestwick, Nicholas J. Wald
• Format: Article
• Language: English
• Published: F1000 Research Ltd 2017-10-01
• Series: F1000Research
• Subjects: Methods for Diagnostic & Therapeutic Studies; Theory & Simulation
• Online Access: https://f1000research.com/articles/6-1896/v1

Background: Prenatal cystic fibrosis (CF) screening is currently based on determining the carrier status of both parents. We propose a new method based only on the analysis of DNA in maternal plasma. Methods: The method relies on the quantitative amplification of the CF gene to determine the percentage of DNA fragments in maternal plasma at targeted CF mutation sites that carry a CF mutation. Computer modelling was carried out to estimate the distributions of these percentages in pregnancies with and without a fetus affected with CF. This was done according to the number of DNA fragments counted and fetal fraction, using the 23 CF mutations recommended by the American College of Medical Genetics for parental carrier testing. Results: The estimated detection rate (sensitivity) is 70% (100% of those detected using the 23 mutations), the false-positive rate 0.002%, and the odds of being affected given a positive screening result 14:1, compared with 70%, 0.12%, and 1:3, respectively, with current prenatal screening based on parental carrier testing. Conclusions: Compared with current screening practice based on parental carrier testing, the proposed method would substantially reduce the number of invasive diagnostic procedures (amniocentesis or chorionic villus sampling) without reducing the CF detection rate. The expected advantages of the proposed method justify carrying out the necessary test development for use in a clinical validation study.