Noninvasive detection of fetal trisomy 21: systematic review and report of quality and outcomes of diagnostic accuracy studies performed between 1997 and 2012.

ABSTRACT

BACKGROUND: Research on noninvasive prenatal testing (NIPT) of fetal trisomy 21 is developing fast. Commercial tests have become available. To provide an up-to-date overview of NIPT of trisomy 21, an evaluation of the methodological quality and outcomes of diagnostic accuracy studies was made. METHODS: We undertook a systematic review of the literature published between 1997 and 2012 after searching PubMed, using MeSH terms 'RNA', 'DNA' and 'Down Syndrome' in combination with 'cell-free fetal (cff) RNA', 'cffDNA', 'trisomy 21' and 'noninvasive prenatal diagnosis' and searching reference lists of reported literature. From 79 abstracts, 16 studies were included as they evaluated the diagnostic accuracy of a molecular technique for NIPT of trisomy 21, and the test sensitivity and specificity were reported. Meta-analysis could not be performed due to the use of six different molecular techniques and different cutoff points. Diagnostic parameters were derived or calculated, and possible bias and applicability were evaluated utilizing the revised tool for Quality Assessment of Diagnostic Accuracy (QUADAS-2). RESULTS: Seven of the included studies were recently published in large cohort studies that examined massively parallel sequencing (MPS), with or without pre-selection of chromosomes, and reported sensitivities between 98.58% [95% confidence interval (CI) 95.9-99.5%] and 100% (95% CI 96-100%) and specificities between 97.95% (95% CI 94.1-99.3%) and 100% (95% CI 99.1-100%). None of these seven large studies had an overall low risk of bias and low concerns regarding applicability. MPS with or without pre-selection of chromosomes exhibits an excellent negative predictive value (100%) in conditions with disease odds from 11500 to 1200. However, positive predictive values were lower, even in high-risk pregnancies (19.7-100%). The other nine cohort studies were too small to give precise estimates (number of trisomy 21 cases ≤25) and were not included in the discussion. CONCLUSIONS: NIPT of trisomy 21 by MPS with or without pre-selection of chromosomes is promising and likely to replace the prenatal serum screening test that is currently combined with nuchal translucency measurement in the first trimester of pregnancy. Before NIPT can be introduced as a screening test in a social insurance health-care system, more evidence is needed from large prospective diagnostic accuracy studies in first trimester pregnancies. Moreover, we believe further assessment, of whether NIPT can be provided in a cost-effective, timely and equitable manner for every pregnant woman, is required.