Count-based size-correction analysis of maternal plasma DNA for improved noninvasive prenatal detection of fetal trisomies 13, 18, and 21.

PURPOSE: Our goal was to derive more sensitive and accurate Z-scores based on combined DNA count- and size-based algorithms to advance molecular diagnostics for noninvasive prenatal testing of fetal trisomies.
METHODS: We included 180 cases at high risk for fetal aneuploidy who underwent amniotic fluid cytogenetic analysis. We calculated their traditional count-based Z-scores, as well as their 100-, 130- and 150-, and 166-bp size-corrected Z-scores, and determined each Z-score's reliability based on its comparison to the cases' cytogenetic results.
RESULTS: We detected for trisomies 13, 18, or 21 among the 180 cases in our study by amniotic testing and DNA sequence analysis. None trisomies 13 was detected, while 1 case of trisomies 18 and 3 cases of trisomies 21 were found. The sensitivity, specificity, and accuracy of traditional count-based Z-scores were 75%, 98.86%, and 98.33%, respectively, while these rates increased to 80%, 99.43%, and 99.44% with a count-based 100- and 166-bp size correction. Moreover, the sensitivity, specificity, and accuracy of count-based Z-scores with 130- and 150-bp size corrections were 100%, and neither of these algorithms yielded false positive trisomies, unlike other evaluated size correction Z-scores.
CONCLUSIONS: Count-based Z-scores with 130- and 150-bp size corrections more robustly predict fetal trisomies than count-based methods alone or those combined with other size-correction cutoffs, such as 166 bp. These testing parameters may enhance the utility of DNA sequence-based methods for noninvasive prenatal detection of fetal trisomies.